Oral History Transcript — Dr. Donald Keck
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
See the catalog record for this interview and search for other interviews in our collection
Donald Keck; November 3, 2004
ABSTRACT: Family background and early influences for a career in physics; early education and hobbies; Michigan State University (1958-1967) B.S. to PhD; physics courses and textbooks; optics research with C. D. Hause, Joe Aubel; Corning, Incorporated (1968-2002); optical transmission research, optical fibers; improving the Fonstad fiber; world’s first low-loss optical fiber, 1970; Frank Hyde’s silica research; leadership and organization at Corning; research at Corning, including glass lasers, radio frequency sputtering, plasma discharge, superconductivity, catalytic converter materials; management responsibilities; university – industry collaboration; Infotonics Center and OIDA (Optoelectronic Industry Development Association); Corning history.
Session I | Session II
Lassman:With Donald Keck on Wednesday, November 3, 2004, continuing the oral history interview. Yesterday, just to go over a few things, we had started with your early career at Corning, in the late ‘60s. We were talking about the decision you made to come work at Corning, the other offers you had received in industry, and also in the CIA. The last question we had talked about was the academic tradition of research that you brought, and your colleagues brought to the labs in the late ‘60s, versus the craft tradition in glass making. You talked a bit about that. The next question I would like to ask is about the opportunities for continuation of your research that you had done in graduate school. What opportunities there were. Did you continue that work or did you immediately shift to projects that were underway in the laboratory?
Keck:The answer is very simple. I shifted, yes. I was hired specifically by Bob Maurer to be a full-time person on this new project that he had started actually a couple of years earlier. Looking to solve the problem of transmission in fiber optics.
Lassman:Can you talk a little bit about Maurer’s background?
Keck:Bob was a physicist — got his PhD at MIT. He came out of the University of Arkansas with his undergraduate work, as I recall, had come directly to Corning from MIT. I don’t know all of the projects on which Bob worked, but one of them involved the ophthalmic lens business that Corning had, and they were continually searching for different glass materials to use in eye lenses, in eyewear lenses. Bob had done a scattering study on different glasses, had measured the intrinsic Rayleigh scattering on a whole bunch of glasses. And I’ve forgotten why he was doing the scattering. But at any rate he wrote a paper. We’ll have to look it up. I was going to say it was published in American Physical Society, but I’m not sure that’s right. But at any rate, it looked at a series of glasses to measure the Rayleigh scattering in those materials. He assembled a scattering photometer and looked at the angular distribution of light coming out of, through a number of possible ophthalmic lens materials. And so, he had a low scattering for eyeglass use. And that paper, I believe, was published circa 1953.
Keck:So, it was early, presumably early in Bob’s career. When did Bob retire? I think he retired in the mid ‘80s, and would have been thirty-five years into the mission. So, what’s that make? Fifty — well in ’53 he would have been fairly early in his career. At any rate, that paper gave him a background to suggest that materials that we look for low-loss fiber optics were in the fused silica family. It was the lowest Rayleigh scattering material that he had uncovered in this earlier scattering work that he had done. He had done a number of projects after that. The group, as I perceived it, had spent most of its recent time on new laser materials. Lasers in the ‘60s, and everybody was trying to find what materials will lase. And Corning was obviously looking at various doped glasses, medium and doped glass in particular, and all of those chromium-doped glasses that I later found buried down in one of the caves under the research building. And so, they obviously looked at a whole series of materials trying to figure out where Corning might participate in the burgeoning laser market. So, a lot of the group was oriented toward that. Mike Teeter was, I mentioned yesterday, a new hire that had come in from the University of Texas, I think, looking at displays, in particular plasma displays.
Lassman:And this was just in the physics department?
Keck:This was just in the physics department was the context in which I came. So, it was largely optics-related stuff. The other groups, led by Sammy Halliby and Andy Herzog, the other people with whom I interviewed were looking at materials for semiconductor use, what glass materials could possibly be more for semiconductor technology. Stanford Ovshinsky had published some of his stuff out of the Ann Arbor area, and amorphous silicon. And so, it was a lot of work in Herzog and Halliby’s department of what some of these new materials might look like. Could we make semiconductor glasses? Herzog looked at materials. Halliby looked at what applications those materials might have. But, Bob’s group was largely in the optics arena, and worked closely with our television division. They were wondering what new technology might supplant the cathode ray tube that was making tons of money for Corning, at the time. So, I think Burroughs was the corporation that came up with the Nixie tube. It was a plasma-based display device. And the question is whether Corning could come up with some variant on that theme, and would it supplant cathode ray tubes. So, two years working on plasma devices.
Lassman:Is this plasma similar to what we see today.
Lassman:I mean this was…
Keck:Glow discharge. No. Well, in the sense that, yes, you’re exciting electronic states in gaseous materials to emit photons. In that sense, yes. But, well I guess — yes it’s probably much the same. Yes. Mike was working on different configurations of small holes etched in glass plates, and could he excite a discharge in the hole and get it to maintain itself. The trick was can you turn it on with a high voltage short-duration pulse and would it say on? Could you get it above threshold, and have it stay on?
Lassman:And, what happened to that?
Keck:Well, he made a few displays but eventually it was decided that that really wasn’t going to capture any of the market. And this was, you know, in the late ‘60s. So, well ahead of its time compared to the plasma displays that we have today. But yeah, the plasma displays today, they’re using different gaseous species, even in colors. Mike was working with neon at the time. But, he had the notion that you could bring in xenon and other gases to get different colors, and so on.
Lassman:So Burroughs comes up with this technology? Were there other companies in addition to Corning that were looking at the plasma technology as a possible replacement for traditional CRT tubes?
Keck:There probably were, but I don’t know.
We had a supply relationship with RCA at the time. I know that there were discussions, meetings between the companies, having to do with the display business, and so on, and my guess is RCA probably had programs in that area as a major laboratory. I have to believe that others did as well. Corning was not nearly as big, so we had one scientist working on this to see if he could come up with a breakthrough. Nick Borrelli was a scientist who, by the way, is still at Corning. He had joined a couple of years before I did. And, as Nick said, the last time I talked to him, he was going turn out the lights. And, Nick was working on nonlinear optical materials, trying to find some glass material that we could make shutters, optical shutters out of, fast shutters. Oh goodness. Miles Vance, who still consults with Corning, had been there a few years. He came from Ohio State, and he was working in lasers, laser glasses. Bob was working on the laser glasses.
But, Bob was the manager of the department, and it was a department of, I think about six scientists and six or seven technical support people in it. But, the size, size of Corning’s department. And Bob had, well, I mentioned Gale Smith as being a person that I had met at Michigan State, as a recruiter. And, Gale had the interesting duty at Corning, in addition to running the analytic services group — a group that specialized in measuring devices and things like that — he was our technical liaison person. And, he basically traveled all around the world. This was in the days when not too many people traveled around. But, Corning realized that not all the research goes on A.) in the United States, or B.) in the Corning laboratory. And so, Gale’s mission was to travel all around Europe, in particular, well and he made trips to the Orient as well, and had built up a whole cadre of friends, acquaintances, that he had met over the years in various laboratories. One of them being British Telecom Labs. And, it was on one of these trips that Gale had made where we learned that they were very interested in the possibility of optical communications; number one, the laser offered possibly the light source, and it had invented and was being developed in the early ‘60s. And, they were looking for some way of transporting that light beam over long distances. And, a lot of experiments had gone on with atmospheric transmission, and you always played with the normal weather maladies that would foreclose that sort of transmission.
So, they were interested in any kind of conduit that could be used to transport light beams over long distances. Gale brought that information back to Corning. There was a scientist at Standard Telecom Labs. And, Gale visited there as well. That scientist was Charles Kao. And Kao had published a paper in 1966 suggesting that he might be able to purify a glass sufficiently well to get the requisite transmission properties in glass to — I don’t recall the paper specifically anymore, but he, somewhere along the line the notion was espoused that if you could get one percent of the light transmitted over a kilometer distance that would be sufficient to allow you to think about an optical telecommunications system. So, that translates into the metric that became the measure of transmission corresponded to the loss rate of twenty decibels per kilometer.
Lassman:So, that was the threshold?
Keck:That became the sort of benchmark, the threshold that everyone in the world was shooting for. And, if you looked at the optical glasses that Bob had looked at back in his day, scattering, he calculated the scattering loss rate of some of those glasses was on the order of four decibels per kilometer. So they would, the scattering, intrinsic scattering, of fused silica, we felt, was low enough to meet the target. The bad news was that if you looked at just the transmission of these glasses, the loss rates were about 1000 decibels per kilometer. And, remember this is a number that’s in the exponent. The loss goes as exponent to the loss rate, times distance; so at 1,000 decibels per kilometer it’s 10 to the minus 1000 power, which is a big number. We had to get down to 10 to the minus 10 times the length, as the loss rate. And so twenty, twenty decibels. Kao had published a paper in ’66. Smith had come back to the Corning laboratory with this information, and had talked with Bob, and wanted Bob’s interest in it a bit. Bob had a couple of part-time scientists in his department looking at the fiber optics that Corning made. I mentioned that Gale Smith had brought a fiber optic to Michigan State when I was a student there. Fibers had really been invented and developed in the early 1950s. Corning had a factory operation, albeit small, to make fiber bundles. Typically, 128 fibers were in the bundles, and these were multi-layered large core, thin-clad fibers. The world of fiber optics had migrated to using them in endoscopes. And before that, you had to maintain the spatial proximity of fibers over three or four foot distances. And Corning never entered that market, and left it to — and I can’t tell you the companies, small companies in New England. Well, Galileo was one of them. Mosaic Fabrications, now that I, my brain has started engaged in some of the history, was another one. American Optical was one that made these coherent bundles.
Lassman:And American Optical was here, right, in Rochester?
Keck:No. It was Sturbridge, Mass.
Bausch & Lomb may have been making them. I’m not sure. But, most of the endoscope work was in the New England, the Boston area. And there were undoubtedly some Japanese companies. Olympus is now the, has eighty percent of the endoscope market. Last I knew, anyway. Corning’s fiber optics were really just light pipes, and they were pursuing the automotive market. At the time, Corvette was interested in having a dashboard indicator as to whether the taillights and the headlights were working. And so the idea was to have a fiber optic bundle that would gather light from your headlights or your taillights and bring them up to the dashboard area, a real-time indicator. And the military was interested in similar sorts of usage, flameouts in jet engines. Could you get fiber optics in a jet engine and telemeter that back to the pilot to tell whether his engine was operating or not. So, that was the direction Corning was going. The fiber optics were a part of the television business. Corning would often have a major business element that was making, well, a few million dollars. Back in those days, as I recall, Corning — I’d have to dig back in the history books — but I think it was about $100 million. I recited the other day the goal was to double it in three years Amo Houghton took over.
So, television was one of the big businesses, and they always had a new business development person in each of the divisions, looking at new areas, projects coming out of the laboratory that could, could be added to that that division’s portfolio. Fiber optics was in the television division, and the new business development person was a fellow by the name of Chuck Lucy. And that name will come back over and over again. And Bob and Chuck had a very good relationship, worked closely together. And, Chuck was very interested in — he was a physicist, a bachelor’s degree physicist that joined the laboratory and then migrated into the business division. So, that was a better career opportunity for him. So, we had the fiber optic small business that Chuck ran as new business development director, and Bob assigned a couple of scientists in his group who weren’t there when I joined. They had moved on. I don’t know whether they were asked to leave, or whether they just decided there were better jobs elsewhere — I think one of them went to Bausch & Lomb, actually.
But, that’s neither here nor there. They had part-time measurement projects to look at the existing fiber optics. The losses of those were on the order of 1,000 db per kilometer, and so the question before the house that they tried to answer, “What’s the source of losses in Corning’s existing fiber optic bundles?” Those fiber optics were using the very best optical glasses that we could lay our hands on. By the way, made by Schott Glass in Germany. Corning had a dickens of a time making optical glasses in those days. We’ll probably have to excise that from the final transcript. But, off the record, if we can do that on a tape. It’s always interesting to me that Corning Glass did not do well in the optical glass arena. Schott Glass was a better melter of optical glasses. And years later, when I was directing a group at Corning I had the opportunity to visit Schott, in Germany. And, the places that they were melting these optical — these beautiful, pure, and clear optical glasses it’s a dirty old — you talk about craftspeople versus academic high technology. And this was definitely craftspeople.
Keck:But, they could melt nice optical glass. I’m mean, to look at these glasses they were absolutely water-white transparent. Both Corning and Schott. But, when you began to really plough into it and make detailed measurements why you found that absorption was taking place by virtue of impurities in these glasses and to a level that would render about 1,000 db per kilometer. Clearly not, you know — orders of magnitude away from the target of twenty that we had to have. But these part-time scientists had found that the losses in these fibers at Corning were about equally divided between absorptive losses and scattering losses. And, together they were about this 1,000 db per kilometer. So, that was the status when Bob began looking around and wanting to hire somebody to work on this project. These two scientists had left the department, and he didn’t have anybody to work on it, and he was looking for someone that could take over a project full time. And, in the door I walked in December, and they made me an offer, and I joined, reported in January of 1968, assigned to work on the optical fiber project.
Lassman:Just to recap here. You described some of the markets that were open for the existing fiber optics business. It sounds like what you’re saying, those are more niche markets. Those are small markets.
Keck:Yes. Relative to what people envisioned in telecommunications.
Lassman:Okay. And then by the time you come in January of ’68, Smith has already done his tour?
Lassman:And the decision’s already made?
He did them regularly. He had brought that information back shortly after Kao’s paper appeared in ’66. So, there were a couple of years and a couple of trips that Smith had made. When I joined, there were two trip reports that Smith had filed about his visits up at British Telecom and Standard Telecommunication Labs, and other laboratories around in Europe, indicating that people were still interested in this. And, that they had active programs going at those laboratories trying to do exactly the same thing. So he had, by that time, clearly convinced Bob that this was a project that he had to assign, had to get some work going on quickly, and try and compete, and get Corning into this, into this market. So, I joined and was assigned to this project. A lot of PhDs when they come out of school, want to continue their thesis work. That was not something that was particularly high on my list. I had done the things that most people do, of identifying the next things that I would do on the thesis, further the work. But it was not something that I was wedded to, or terribly interested in continuing. So when I was assigned the project, it was about wave guidance. I like electromagnetics.
It was about optics. I liked optics. I didn’t know anything about materials. And it just seemed like it was an interesting area to look at. You had asked a little bit about the climate in the laboratory at the time, and I clearly came to Corning believing that this was the laboratory that was going to do world-leading work. And it was, the scientists that I talked to, the managers impressed me with the notion that they hired the very best. MIT was, a number of the people that I had talked to had come from MIT, including Bob. They valued publications, which sounded good to me. And implied to me that they were one of the major laboratories in the world, and indeed they are, were and are. And so, the academic sort of work, it looked to me that I wasn’t foreclosing too much on the academic type of career. In fact, one of the questions I asked was were there opportunities to teach, part-time, or so on, in the area? And, the answer was that a number of the scientists, in fact, were teaching in some of the schools nearby. Elmira College being the one that a couple of the people were teaching at. It looked to me as I was coming into a very high technology company, and I wasn’t even aware of the crafts aspect of the thing. Obviously, very quickly, I got that education. But again, it didn’t bother me.
At Michigan State, we had craftspeople who were in the physics department. There was a machine shop, and I had worked with them, and gotten to know the fellows, and as I say, had no problems bridging between the talented craftspeople and the scientists with whom I interacted. So, at any rate, I joined [Corning] in January, and found that there was a development group that was interested in the project and was assigning sundry sources to it, albeit part-time. There was a — oh shoot. I’m not going to remember the guy’s name now. Well, I don’t remember the manager of the department. The gentleman that was working for him was by the name of Frank Zimar — Z-I-M-A-R. And, he had a fiber-making operation, a high-temperature fiber-making operation in the development lab. The factory was downtown in one of the buildings. Bob had hired a summer student, Cliff Fonstad – Fonstad, who is a professor at MIT, now. Still there I believe. The summer before I came, they had made a fiber out of the high silica glasses.
Lassman:Are these the glasses that he had published in that paper?
Yes. But I need to maybe give you a little background on a fiber. As fibers in that day were being fabricated, there was a core material and a cladding material. Very early fibers were unclad. They just drew a long strand of glass, and because air was a lower refractive index than the glass, light would be totally internally reflected at the glass-air interface. But any perturbation and touching of that glass would cause scattering to take place, and so on. Eventually the industry moved to these so-called clad fibers. You needed two dissimilar materials. One a higher refractive index material, and then a surrounding low refractive. Well, fused silica, at the time, had the lowest refractive index that we knew about in the glass. The question was, how are you going to make a fiber out of fused silica. And yet, Bob had the notion that this was really the material that was going to have the best shot at making a low-loss fiber.
So, the notion that he came up with, and this was Bob’s and it’s the contrarian notion, fascinating contrarian notion. His notion was, “I’ve got to raise the refractive index. How do I do that? I add an impurity to fused silica and the index increases.” Generally, if you add an impurity to most any material with a few notable exceptions, the refractive index in the material will increase. So also will your Rayleigh scatter; well and perhaps also your absorptive losses. The obvious question that anybody outside of Corning was asking was, “Well, why would you add an impurity? This is exactly in the wrong direction. Right?” Bob had that as the basic premise. Now, Corning was making, was manufacturing, had been for a long time, manufacturing a titanium-doped fused silica. It terms out, if you put about seven percent titanium oxide in SiO2, fused silica, you get a material that essentially has a zero thermo expansion coefficient around the room temperature plus or minus a 100º range. So, Corning had invented this material as a possible telescope material. I have a wonderful story on that, by the way. I don’t know if you’ve read the book, The Perfect Machine, but if you haven’t I will commend it to you.
Keck:I’ll get you a copy, or show you a copy later on.
Keck:Tell you about it, or whatever, when we go to lunch. So, Corning had been making this for quite some time. And, it raised the refractive index of fused silica. So, his notion was we put titanium in, titanium-doped silica as the core, and we’ll clad the fiber with fused silica. And, we knew —, they speculated that both materials were going to have to be very very low-loss. After I got there and began to look into the propagation theories and look at where the light was distributed in these fibers. You realized that only about eighty percent of the power was actually carried in the core. Twenty percent of the light was, in fact, traveling in the cladding as an evanescent wave. And so that cladding had to be low-loss, otherwise it would siphon away energy away from the wave as it propagates. So, that was the design that Bob and Cliff Fonstad had tried in the summer. They just titanium-doped the core, seven percent titanium, and clad it with silica. So, this fiber was waiting for me when I got there. I was to set up measurement equipment and ascertain what the losses were. There was a whole host of other things that we knew we had to do. But, one of the things that we knew about titanium doped silica was at high temperature titanium goes into a reduced state. Instead of TiO2, which would stoichiometrically fit the silica matrix it goes into a TiO state. And so, you wound up with a non-bridging oxygen atom in the matrix. And, a lot of them. Well that non-bridging oxygen is sitting there with a dangling electron and absorbed light in the blue-green region of the spectrum. And the tail, eventually, we found that the tail of that band was to extend way out into the infrared, and absorbed any light that we tried to transmit through it. So, when I first plugged these things in the laser rig that I set up, and I had a small helium neon laser there, and I set up equipment to measure the transport properties, and rigged up a little microscope so that I could focus the laser beam on the end of the fiber, and with a beam splitter actually see where the spot was on the end of the optical fiber. This will become important later on.
Keck:And, slid it over the core of the fiber, and you couldn’t see the core. The glasses were sufficiently of the same refractive image. There was no change in reflectivity, but you‘d get the spot somewhere near the middle of the thing, and then look at the other end of the fiber and see if there was any light coming out. And generally, there wasn’t a peep of light coming out. I’d cut it off, and cut it off, and cut off, until I could see some light coming through and then start measuring the loss rate as I chopped off pieces of fiber. Well, we were looking at pieces of fiber a few centimeters long; was all that you could transmit through. And I said, “This isn’t going anywhere.” Well, we knew that the titanium was in this reduced state, and other scientists, glass scientists around Corning knew that if they heat treated the titanium-doped silica they could re-oxidize the state and get that dangling oxygen bond tied to something, and get rid of this absorption band that was occurring. So, the first set of experiments that I ran were to heat-treat pieces of this fiber. And, as I heat-treated it, indeed the absorptive losses came down, and down, and down. At some point, when we were heat treating it, at well, 600 to 1000º Centigrade, most optical glasses would long since have melted and run all over. But, fused silica doesn’t melt until you get it up to about 1100º Centigrade. But, well, lots of aspects to this story. As I heat-treated it, why indeed the losses came down, and down, and down, and then they would turn around and go back up. I mean, my job was to figure out — well, we knew why they were coming down. We were oxidizing the titanium, getting it in the right oxidation state. So, I tried to get the analytic services group to do electron spin resonance and NMR measurements of that glass to ascertain what oxidation state the titanium was in. Had we oxidized it all, or hadn’t we? Why did it turn around? And then also with my laser I was finding that the reason it was turning around was that I was growing scattering centers. Great big…
Lassman:While you were oxidizing?
Keck:While I was heat-treating this thing. I’m sure there were bubbles in the glass so small that initially they didn’t cause any problem, because you heat treated it, why these bubbles would grow and they would crack the glass, or whatever. And, they tended to occur at the core-cladding interface of the fiber -– and I could see these. And when you had, when you heat treated it and gotten it up very high, why there would just be these great growing clunkers, and as I used to call them, in these short pieces of fiber. So, I could zero in with a microscope and actually see it, there was a damage at the core-cladding boundary. Well, the other thing that was a problem with these materials is, and I began at that point learning a heck of a lot about glass. Fascinating material.
Lassman:I was just …
Keck:I knew nothing about it coming in.
Lassman:I wanted to ask you about that. What was the learning curve like?
Lassman:How did you negotiate that, that learning curve, or keep up with it?
Keck:I was able, as I described, in my growing up, I could bridge the space between groups of any kind. So, I talked with the craftspeople on the shop floor that were growing fibers. I talked with the people, there was a melting group. I talked with them. Other scientists that worked as chemists, or process people in the glass industry. And so I had, I guess I probably, in retrospect, pride myself on having connections all around the laboratory. And, I just, I knew…
Lassman:You were just talking about the learning curve, working in the glass technology.
Keck:Yes. So, I was able to and doggedly made my rounds around the laboratory, and learned who the important scientists were that had a right to an opinion, and whose I didn’t need to trust. And so, gradually [I] came up the learning curve on the glasses. One of the first things we found as I heat-treated these pieces of silica was that at those temperatures, and we were treating in an oxygen atmosphere, under those conditions, you grow silicon crystals. I shouldn’t say silicon. Silicon dioxide crystals, instead of an amorphous material, there you suddenly get crystal formation taking place. Well, these crystal have a different set of properties, mechanical properties than the amorphous glass around them. So, these fibers would come out, and if you looked at them crossways, they’d break, just fragile as all get out. And, the longer I heat-treated it the more crystoballite I grew on the outside of the fibers. So, these things. And, the fibers, I was heating treating them in an oven on a Vycor spool. I had to find a material that would withstand 1,000 degrees C. So, I made a reel. My technician, Larry Carpenter, and I made a reel out of Vycor — which is basically fused silica — and had glass rods between two plates. And so, we could wind our fiber on this drum, and put it in the furnace. And then I’d just take this out and put it in a laser rig and carefully try and pull off one end, get to my microscope and my laser beam, and then gingerly take the other end and put it in my detector, or what have you. Well, so these fibers were extremely fragile. And, I can’t tell you how frustrating it was to, you know, get a nice fiber and almost get it set in your rig and it would break. And, I was trying to cut back as I unwrapped it from the reel. And, if it broke too close to the end then I’d have to start all over again. It was an interesting learning experience to figure out how to deal with this glass.
Lassman:Did you ever think…?
Keck:Interestingly I never, at that point, thought about, well, even if I get the transmission up, how the heck was anyone going to use this? But that came later.
Lassman:I was wondering. Was there ever a point where you thought this fused silica just isn’t …?
Lassman:The mode of operation?
Keck:No. I suspect from my grandfather, I probably had this notion of just perseverance. And, there was never a question in my mind that what we were going to succeed. Now, about two years into the mission the company started having second thoughts about our spending. Shouldn’t we have this scientist looking at some other problem?
Lassman:This is around 1970?
Yes. At any rate, the first set of experiments involved looking at this Fonstad fiber in heat-treating it. I think I filed my invention disclosure in February, and I — well, leave it at that for the moment. I was trying to recall something else. We, of course, knew Fonstad’s fiber wasn’t the answer at this time. We knew we’d have to make another fiber. I recall that there’s a group that got together and at this point in time, I was introduced to Peter Schulz. Peter was a glass chemist -– actually a ceramist from Rutgers. He graduated from Rutgers, and joined Corning about six months before I did. And he joined the glass research department. Bill Dunbar was his supervisor. And, Bill and Bob Maurer were on the same level in the organization.
Bob reported to Chuck Wakeman. Bill Dunbar reported to John McDowell, the director of chemical research. Pete was hired to investigate the vapor deposition fabrication of glasses. All the things you eventually learned at Corning was that, remember the ideas and the technologies all along because they’ll come back. Frank Hyde was a scientist in the 1930s, I think, [when] he joined Corning. And, a brilliant scientist. He is in the National Inventor’s Hall of Fame, as well, as we are. And, he had, this is where “The Perfect Machine” comes in, Frank had come up with a way of vapor depositing fused silica, in particular, starting with raw chemical materials, silicon tetrachloride, eventually titanium tetrachloride, and if you burn those in the presence of oxygen the chlorine vaporizes, the oxygen combines with the silicon, or the titanium, and you get little particles of sand.
The beauty of the process that we really found out as we pursued our fiber work, the beauty of the process was that there’s an automatic distillation that takes place. And, all the stuff, the impurities, iron, chlorides, all the rare earth chlorides, that might come through in the silicon tetrachloride batch have a lower vapor pressure than does silicon tetrachloride at the temperature that you operate the processes. And so we basically distilled the way (or away) the typical impurities in glass, iron, copper, cobalt, that give color to glasses. And so Pete was hired to — we already could make titanium. Hyde had come up with that. And we could make silica that way. And, in fact, we had a factory in both Bradford, Pennsylvania, and then Canton, New York. to make these materials. And they were making boules of the stuff that were ten, twelve feet in diameter, a foot thick. Huge pieces of these materials. And, I’ll tell you more about that as we go on. So, that sprang out of the Hyde work in the 1930s and then they built the factory. And Hyde, by the way, invented that process to make telescope mirrors, ala the 200-inch telescope, which was how it ties to “The Perfect Machine.”
Keck:The other thing that Hyde invented — he had stopped that work and went on to other areas. He had the notion that all glasses, most glasses, involved oxide materials, oxygen in the matrix — some cation, and oxygen — and he had the notion, “Well, what if we replaced the oxygen with carbon?” He began looking at silicon-carbon compounds — a whole family of organic materials. And, in fact, invented silicone, in the Corning laboratory. We didn’t know what to do with it. We’re not an organic company. We don’t sell chemicals. And, the senior Houghton, Jamie and Amo’s father, knew the president of Dow Chemical. They got together and at the highest level of the corporation formed the joint venture. Dow-Corning. And, Hyde moved to Dow-Corning and became a chief scientist at Dow-Corning. He came back to — I might have met him once, or twice, during his lifetime. He worked with Pete and the chemical folks most of the time. But at any rate, Pete’s job was to see what other materials he could make, presumably for deposition process. Well, circa February of 1968, I don’t know whether Bob convened the meeting, or whether a development guy convened, or who convened it, but there was a meeting of Frank Zimar and his manager, who was, could make a silica fiber; Dunbar and Schulz, and Maurer and Miles Vance was kind of mentoring me. Miles was doing his laser project, but they kind of had him looking over my shoulder and making sure I was doing the right things and getting the things I needed, and so on. And, so a room full of people and we discussed how the heck we were going to make this optical fiber. The problem was we were trying to make a single-mode fiber from the get go. The question was, how do you make a very very small core and a great big fat cladding? It was the evanescent tail and the way it was going to extend out to infinity and we defined infinity being at least 100 times the wavelength of light out on the cladding.
Lassman:So, that’s why you need the thick cladding?
Keck:Yes. That’s why you need the thick cladding.
Lassman:Okay. Just to make sure I understand.
And the ways that Fonstad made his fiber were the tried and true ways that everybody was making fiber in those days, the so-called rod-in-tube method. We took a rod of the core material, and you put it into the tube of the cladding material. Now, the problem was if you make a rod, it’s going to have finite size to it, maybe a few millimeters in diameter, if you draw a cane of glass. And, now how are you going to make this great big thick cladding, drill a hole in the stuff and put this rod in the middle of it, and draw the whole structure? The only way we could see coming out of this meeting was if we made two draws, where we put the titanium-doped rod in a tube of silica and then draw that into a rod, and put that rod into another piece of silicon, and draw it again. And then we get two interfaces and they hemmed and hawed about each time we had an interface that’s a problem, and so on, and so on, and so on.
Then I went back to the laboratory and came up with this notion based on my interview trip at IBM, where they told me about radio frequency sputtering and being able to make thin films of amorphous materials, or dielectric materials more properly. And, I had met John Kerr, a fellow up in Sam Halliby’s group who was trying to learn how to make glass materials by R-sputtering. So, they had a sputtering rig up there. He was doing it in planer geometry, because he was trying to make something a la Stanford Ovshinsky and wafers of amorphous silicon for the semiconductor industry. But, he knew about sputtering glasses and so on. Well, I said, “Gee that, that’s the way to do this. We want a tight bond between the core and the cladding.” And the rod and tube, the problem of course is that you’ve got this area around the rod that you stuck in the tube. You know, have you cleaned it properly? Is the inside of the glass tube clean? How do you ensure that? Are you putting it in this high temperature furnace, all sorts of debris floating around in this high temperature furnace. Materials falling off from the muffle that is used in the furnace, generally high temperature ceramic materials. So, there’s all sorts of refractive dust floating around. And, we were speculating that a lot of this dust was getting into the interface between the core and cladding on this rod too.
Later on, in fact, I set up some scattering, angular scattering measurements in the fiber and ascertained that some of these scattering centers, in fact, were me scattering, not Rayleigh scattering. So, definitely on the order of wavelength or, well, bigger than wavelength, from the very forward directed. And so, we were pretty certain we were getting some of this refractive dust in the core-cladding interface. So, the question was two-fold. How do we get the large core-cladding ratio that you need for the electromagnetic properties, and how do you get that interface to be pristine? So, the notion I came up with was, well, let’s deposited a thin film on the inside of the tube. That small core volume, as we make thin film as thin as you wanted it, and you deposited it intimately on the inside wall of the tube you were starting with. Then the only problem was how you keep stuff from going into the hole in the middle. Well, we figured we could probably plug up the hole and clean the, make sure the thing was clean before we put it in the furnace and then draw the fiber down while it was plugged up and be able to keep the debris and stuff from getting inside.
Lassman:But when you insert the film between the rod and the cladding don’t you introduce another problem with…?
Keck:Well, I was going to do it, I was going to radio frequency sputter it so it’s all done in a vacuum. And I would have cleaned out — and John Kerr — the other aspects of the sputtering process, you strike a plasma in order to do the sputtering process. You could do plasma cleaning of the surface. I could clean the — I thought I could perhaps clean the inside of the tube before I sputtered the film on it. You do it by the plasma that you generate to actually do the sputtering.
Lassman:You wouldn’t have to worry about additional scattering or loss because of that film?
Keck:That was the intention. That was the speculation. I filed my first invention disclosure in, I think, late February 1968, coming out of that meeting. And I remembered posing the idea to Miles Vance. And, he said, “Oh no. It’ll never work.”
Lassman:This was just a couple of months after you were there?
Keck:Yes. I’m solving problems. Somehow, my brain was attuned to doing that. How do you get out of the box? How do you — I don’t know. I can’t describe it. But, the record will show that the invention disclosure was filed in February of ’68. And, Miles, as I say my mentor, advised me, “Don’t even bother filing a disclosure. They won’t look at it, and it’s not a good idea,” and he cited all the reasons why it wouldn’t work, and so on, and so on. I don’t remember — I think I went ahead and filed it anyway, and the development guy, Wes Jenkins, he talked to Bob, and I happened to be in the secretary’s office while he was talking through the open door. And he’s thinking this is the greatest idea he’s heard of.
Lassman:And he’s from…?
He’s in our development group. He was in — Frank Zimar had the high-temperature fiber draw furnace. I don’t remember what other project Les had, but we were in the fundamental research area, and Les was in a separate building where they did all the development activities once a feasibility had been demonstrated on a project. Went over to the development group — and they picked it up and further refined it, and tried to figure out how you’re going to manufacture it. Wes liked my idea and Bob authorized me to set up, or try and set up in our RF sputtering group. And later on in some of the deposition, I don’t know if I’ll get to that or not, but I never — I did set up a sputtering rig. How do you get something on the inside of a tube? So, I arranged to have the cathode of my RF sputtering rig coated with a Vycor tube and I put that inside the tube. Now when I put my RF potential between the metal rod and the middle I assumed that, and the plasma was going to take place between the Vycor and the silica tube that we were trying to deposit on I was just using Vycor as a temporary trial. But, the plasma would be struck between the, in that intervening space, and material would come off the Vycor.
The micron plane was out somewhere around the outside of the silica tube. And Mike Teeter knew about plasmas, and so he helped me get the plasma going. Indeed I rigged up a vacuum system, had gotten an argon plasma going in the interstices and I think I sputtered some Vycor, but the films were splotchy and colored, and I don’t know why. We didn’t ever figure out why. But, Pete Schulz, by this time, had actually made some glasses, different materials. He put tin in, and cobalt, and any number of other materials besides titanium. And, interestingly he could never, he tried germanium and it didn’t work. There’s a story associated with that. But, Pete, then posed the thought of, “Well, what if we arranged to use the flame hydrolysis process that spewing out this high purity silica particles, or doped-silica particles, and we simply direct it to the hole in the middle of the silica tube and let the soot deposit on the inside wall of the tube?” “If we could get our thin film that way.” And so, in parallel with trying to get this sputtering thing going, I said, “Yeah. Let’s give your thing a try.” Well, Pete had all his vapor depositing stuff up on the fifth floor. And the hood had burners, special burners that had been invented for the Canton plant to deposit the titanium-doped silica. And you had to have oxygen curtains and all sorts of — it was a fairly sophisticated burner. So, we had to do the experiment up in Pete’s lab, if we were going to do it. And one of the other things that I was working on at the time, we needed the silica tube, and we needed it to have a very thick wall.
So, I guess we’re probably four months into the mission now. We’re mid summer of ’68, approaching fall. Pete was called down to go up to the Canton plant and troubleshoot a problem with these huge boules of silica. They were getting all sorts of impurities in the glass that no one knew their origin and the customer didn’t like that. Pete had to go up, and I went with him. One of the sidelights was I was to purchase one of the best boules to get the silica material we were going to use in making those tubes. We went up and my first visit to a factory. Fascinating. They had row upon row of these huge furnaces, multiple burners in the top of this dome structure, laying down these ten, twelve foot diameter boules of titanium-doped silica, really thick. The temperature of the deposition was 1500 degrees C. Fascinating process. But, at any rate, as they, and this was on a big turntable, this twelve foot, ten, twelve foot diameter boule is rotating around. It must have had ten, twelve furnaces each making these boules, different stages. It took days to lay down the layer upon layer upon layer of these high purity silica and build up the thickness of the boule. Fascinating.
Well, at any rate, I visited the plant with Pete, and we got our boule. We looked for an ultraviolet grade silica. And one of the interesting things that we found, there were strata in the silica each time the basic problem with silica is if it cools down its refractive index properties are a function of its temperature profile. As the flame is coming around and depositing the material the thing rotates, so it’s hot right under the burner. But right adjacent to it, before the next burner comes along, it has a chance to cool a little bit. Well, in the act of cooling it changes its refractive index. And then sets. And the next burner comes along and isn’t hot enough and nor do you want it hot enough that it re-melts the underlying material. You want that material to stay there so you can lay the next layer on top of it. So you wound up with these refractive index layers in the material. Later we found if you core drilled a plug out of these foot-thick boules and you had the strata this way, when you put it in the furnace, low and behold the viscosity properties were different.
So, you had gobs of stiff glass and then there’d be the fluid glass and then stiff glass, and then fluid glass. And we found you had to core drill parallel to the strata. But it took us one try to learn that. So, I came back with this boule of silica and took it down to the crafts shop and they were to core drill a plug of silica out of these things, about, the biggest core drill they had was about an inch in diameter. And I had them make a quarter-inch diameter hole exactly in the center. And darned if they do it. I mean they, and how do you — I mean, these are hard materials diamond silicon carbide grinding rigs and drill press and you’re trying to drill a straight hole through a hunk of glass six inches long. But, they did. But…?
Lassman:I just want to ask a question on this.
Lassman:You mentioned the gradient, the difference then with the cool — there’s no way that you can actually reheat the finished piece in a way that will…?
Keck:It’s just, it’s just too high temperature, too refractive a material.
Keck:You’d have to take it up to the softening point, which would be 1500 degrees Celsius in order to get the flow, and if you get it to flowing, why then you’re going to lose the structural integrity of the thing. You couldn’t get it up to that temperature.
Lassman:So, that’s just one of the limitations of the material?
Keck:It is. And I think it is, and it may still be to this day. I confess I — we still have the Canton plant going. And that was one of the problems that GE ran into in “The Perfect Machine” by the way. GE had the notion that they were going to plasma spray silica about the same time that Hyde was doing his flame hydrolysis. GE was, had the contract to make the 200-inch telescope. Hale had given them a $6 million contract to make this thing. And, I’ve forgotten now the well-known scientist of that era at GE who was running the laboratory had…
Keck:No. It wasn’t Langmuir; can look it up. It’s in “The Perfect Machine.” At any rate, his notion was that they’d put sand on the surface of the mirror and then take a plasma torch and fuse it in, melt the sand, as pure a sand as they could find, and build up the layer upon layer for the mirror. And they never could, just couldn’t get the temperature up. You know, it was always the point, a point source. So, they wound up with all sorts of undulations, and variations and homogeneity of the material, and therefore the expansion of the material, and so on, and so on.
Lassman:Okay. I see.
Keck:And, Corning, when Houghton met with Hale in the, in one of the clubs in New York City, can’t tell you which one now. It’s in “The Perfect Machine.” After GE had been trying for three years, and they had great publicity, about making the 200-inch, the world’s largest telescope, Houghton arranged for a $600,000 contract with Corning to try and melt it out of more meltable glasses. And it had already been an experiment run — George McCauley, I think, was the guy’s name, at Corning, mid-1920s. Yes. Mid-1920s had already made a sixty-inch mirror by pouring a boule of silicate glass, basically Pyrex but some variation on that theme. So, Houghton knew he could make something that big. He knew he could make something 200 inches. And, told Hale he could, took the contract, and came back and two pours. We’ve got the original, at the museum, that cracked, and the second one was the one he made. But GE had all sorts of publicity had gone on at the time.
Lassman:Interesting to get to hear about the limitations though.
Keck:It was a beautiful story. A beautiful story. And the problems that they went through in the annealing, and a flood came in ’34 while they were annealing in the Corning factory, the lowest part in the factory. They were jack hammering the transformers that were driving the annealing furnace. They had to anneal it for what, a year and a half, two years, something like that?
Lassman:It took a long time.
Keck:And the water’s rising, as these guys are jack hammering in the midst of, you know, high voltage transformers. They’re driving this furnace and they’re trying to…
Lassman:Can’t be any more dangerous than that.
Keck:But interesting. So, now I lost my train on the story.
Lassman:We’re coming back from a break now. You were just talking about the boules — is that the correct pronunciation?
Lassman:The two limitations. One is the viscosity problem, and then the index of refraction changes as its cooling?
Lassman:And so, at that point you have to…?
Well, so, we eventually figured out how to take this ultraviolet grade silica boule that I got from the Canton plant, and properly drill a cladding tube out of it, and basically had to drill parallel to the strata that were in the silica tube. But, we were then confronted with the question of, after the craftspeople had very nicely core-drilled this plug, about six-inches-long plug of silica, about six inches long and maybe an inch to an inch and a quarter in diameter, with a quarter inch hole through the center of it, how were we going to clean up the debris and imperfections, the cracks, that had come about as part of the core drilling and the grinding process. We investigated a number of different ways of doing that. And, a technician rigged up just a polishing operation, where it was carefully polished with ever finer and finer abrasives, polished the inside of the tube. We didn’t much care about the outside. I mean, that would heal as we, as we drew the fiber, but the inside had to be, we figured, absolutely pristine. And so, he would polish the inside of these tubes. But, part of the problem there was, you’re using abrasives, like cerium oxide, and cerium is an absorbent impurity so we didn’t want that in the center, we knew that.
As he polished it, we’d periodically dip it in HF to get rid of the abrasives that were in there. And as soon as you do that, very often you uncover the micro cracks that the grinding and polishing process creates in the underlying layers of glass. So, you know, you’d roughen the surface up and then you’d grind some more and you’d roughen the surface. And so, that didn’t look like it was getting anywhere. And, I came up with this notion of rigging up a small lathe and in the chuck of the lathe having a, you know, a steel rod is what I used. You put a, polished a 45º angle on the end of the rod. And we got a CO2 laser. We just bought one over in our process research center. And, figured it had enough power if we focused the beam we could probably melt something, like silica. And so then the idea I had was that we’d shine this CO2 laser beam off this rotating 45º mirror that I put on the end of steel rod, and we’d run the cladding tube up and down on the bed of the lathe as the CO2 laser beam is rotating around on the inside of the tube. And, we would arrange the lenses and the distances such that the CO2 laser beam was going to focus right on the surface, inside surface of the cladding tube. And this mirror’s spinning around. And so, the beam was going to describe a circular pattern on the inside.
Well, indeed it melted the glass, and but it’s the same story as with the boules in the Canton plant of locally heating silica. You locally heated it and that would heal nicely but as soon as there was an adjacent spot, why it wouldn’t heal and [we] kept having to try and adjust the rate of traverse of the cladding tube on this thing so we could hit every spot on the inside of the tube. I drew, I made many screw threads on the inside surface of the silicate tube trying to get that just right. And always wound up with some ridges, some variation on the inside surface. So, the laser technique didn’t work. And, Frank Zimar was the one that came up with the solution. He took the idea I had for running a laser beam up inside the tube instead of using laser, he simply got an oxy-hydrogen torch and we made an oxy-hydrogen torch on a small glass tip, and he simply ran that torch through the inside of the tube.
The hydrogen, oxy-hydrogen burners spewing out its heat and we just ran that through the tube and it just healed the inside surface of the tube beautifully. Now we had our tube and that’s what this picture is, by the way, is the laser kit. I’m showing Tom the picture on the set up. And, I can get you a copy of the photo. It’s not proprietary. [Recording paused.] So, after Frank Zimar came up with the hydrogen torch to polish the inside of the tube then the trick was how do we get the tube to rotate up in front of Pete’s, Pete Schulz’s burner, up in his laboratory on the fifth floor? And so, we built a lathe out of a ball bearing. Went down to Elmira to one of the big mills down there, or mill supply companies and bought a ball bearing with a two-inch diameter throat, and mounted a stainless steel tube in the middle of that and drilled some screw holes in it, at quartering (or quarter-inch) positions on each end, and this became our chuck. And, the screws and clamped the fused silica tube that we had flame-polished. We rigged up that small motor to rotate this ball bearing.
The blank, the cladding blank would rotate, put it on a cart and wheeled it up to Pete’s fifth floor lab. And, I have forgotten now the date. I can find it out of the Corning data books as to when we made the first attempt at a thin film inside the tube to fabricate a fiber, but it was probably late ’68 or early ’69 that we made this. Took it up to Pete’s lab and the first, first attempt, the hole in the middle of the blank was only one-quarter inch in diameter and wheeled it up to Pete’s lathe, blocked the burner, and figured we’d have to time the deposition process because we just wanted a thin soot stream inside of the tube. And, we removed the baffle and soot headed for the tube and just coated the end of the tube. None of it went down the middle. Now, what are we going to do? And, I don’t remember any longer whose, who spotted it, or whose idea it was, but there was a vacuum cleaner, an old canister vacuum cleaner in the laboratory that they used to clean up whatever. And, one of us grabbed that, cleaned off the end of the cladding tube, and got rid of the soot that we’d done in the first pass, blocked the flame again, set this vacuum cleaner on the output of the back end of our rotating cladding…
Lassman:And you sucked it through?
Moved the burner and the soot went through that middle of the tube just slicker than a whistle. So then the quandary that we’re in — this is a flaky porous soot that you’re depositing. How do you know when you got the right thickness to make the right diameter core? And so we eventually rigged up, after the first or second pass, we’d routinely put a little tab on the inside of the tube to be removed so we could look into the microscope and measure the thickness of the soot layer. And Pete had done some crude calculations — not calculations — experiments that showed what the densification ratio of these, of the soot materials was. And, as I recall it was like a 7:1 ratio that the soot would densify to about 1/7 of the original thickness. Snow will densify to 1/8 of its thickness and to a water layer, in case you’re interested. So, at any rate we measured that and so we got some idea of then how long a time we had to leave the burner blowing soot through before we shut the burner off and then I’d, the deal was we’d wheel our rig out of Pete’s lab and I’d take the blank over to Frank Zimar and put it into the redraw furnace and draw fiber out of it. And then we would, as the first fiber came out, why I’d quickly take the fiber, run back over to my lab and we’d get an actual diameter measurement… The diameter of the fiber was in wavelengths, to be single-moded, and helium neon. And so then, I’d come running back to Frank and say, “Well, draw the fiber at this diameter and that’ll get us a single-mode operation.” We weren’t sure whether that would be uniform along the whole blank.
So, typically I’d, I’d have say, “Frank, the average diameter you want to shoot for is this, but let’s bracket it, and draw some at this diameter, and this diameter, and this diameter, and this diameter.” So, we’d simply vary the outside diameter of the fiber in such a way that we knew that, we figured we had a pretty good chance of getting a single-mode operation. And then, I’d take the fiber — the first fibers were still made using titanium-doped silica. But instead of the seven percent that we were doping, Fonstad’s original fiber was, the reason that we wanted less titanium, we didn’t need the large refractive index that you had in ULE glass. So, we would reduce the amount of titanium that Pete put in as a doping agent, hoping we get less absorption, number one, and well, period. It was a smaller absorption point we were shooting for. So then, I’d bring the fiber. We’d wind the fiber on big cardboard drums. This is an interesting piece of history, or at least I find it so.
These cardboard drums were reject or left over from our melting operation. We’d get raw materials, melting, for our experimental melting, in great cardboard drums, maybe a foot and a half, two feet in diameter, a couple of feet long. And, Frank Zimar had a rig where he could level-wind, like a fishing reel, level-wind the fiber on these drums. And we would try to make a single layer. If the fiber, this was uncoated fiber, so if it, if glass came in contact with glass it would break and you get a breakage. So, we knew that after — well, we knew that. Zimar knew that. We’d level-wind and tried to get as big a diameter as we could. I’d take this drum back to the laboratory and we’d typically have several drums at different diameters. We’d try and see if there was any transmission. Typically, there wasn’t. So, we’d have to go through the heat treatment cycle again, and figure out where we were. Learned a lot about electromagnetic propagation. Took some, I thought, beautiful pictures of mode patterns in fibers. We’d actually see the mode structure and we’d use different polarized inputs to excite different modes in the optic fiber, and dielectric waveguide modes. And so that was, that was fascinating.
That indeed proved that we had the right diameter in where we were in the operating curve of wave propagation. But, we were still by-and-large seeing the same sorts of behavior that, if you heat treated it at the right time and temperature, you could come down the curve and get low-loss at some minimum point. I’ve forgotten. We had dropped from 1,000 db per kilometer in some of these down to on the order of 100. So, we had definitely made progress. But, if you heat-treated it too long you’d grow crystals and the curve would turn around again. We largely felt that it was just that we hadn’t perfected our techniques enough as we were doing these things. And we were allowing some dirt in, or didn’t have, maybe could reduce the titanium concentration a little bit more, and so on. And the typical sequence was that I’d make a fiber about every quarter. It took us about three months to analyze what we had done the time before and figure out where the operating point was and what the time, temperature, heat treatment needed to be, get other measurements made on the fiber of the actual titanium concentration that, we were getting what Pete thought he’d put in? We had the analytic services group making those measurements to tell us where we were. And, I, at this point, I’d have to look back in the laboratory notebooks to tell you how many of these thin-film deposited fibers were made before the one that achieved our results. But, circa 1970, in — I’ll have to look back. I can’t tell you the dates anymore. But, 1970 was the year in which we fabricated the low-loss fiber.
We made one that my measurement showed had achieved our goal before 1970, probably late ’69 I’ll say at the moment, but I’ll have to check that, and we can here. I can do that later on for you. But, it was a three-month cycle, and I had gotten Pete to make our blank, drew the fiber, and in the process of drawing something went wrong in Zimar’s furnace. We drew only twenty or thirty meters of fiber. The fiber contacted the sidewall of the furnace and stuck; ruined the blank, the muffle, the whole thing. We had to shut the furnace down and rebuild it. Had this thirty-meters of glass, took it back to the lab, measured, and heat-treated it, measured it, and it was below 20 db per kilometer. The problem is that in thirty meters the change in transmission that would correspond to a loss of one percent over a kilometer, was within my measurement on certainty; I just couldn’t see any change in the light bubble over that thirty meters. It was just too short a distance, twenty or thirty meters. I’ve forgot now. But at any rate, it was within my measurement and certainly and all I could say was that in within my measurement certainly it looks as though it is less than some number. And the number was twenty. I’ve forgotten exactly what it was. So, we thought we had it, but obviously, we just couldn’t prove it. In fact, the next attempt everything went well. We drew a kilometer of fiber, I-100, was the designation for the draw, and my technician and I were beginning to realize at that point that we were going to have to heat treat a substantial amount of fiber.
So we wound, I think, 500 meters of fiber on this Vycor drum to heat-treat, and heat-treated it, and it came out of the furnace and it broke. And, we managed to piece, pull the two ends out — I might have rewrapped that. Well, I didn’t, but Larry might have rewrapped that on another drum, same diameter because it had this permanent set in it. When we heat-treated it, it deformed to fit that diameter, so it would always come out with a curvature to it. And I’m pretty sure that the thing broke, but we were able to salvage 228, twenty-five meters, of fiber in one piece. And, measured that and it registered about 16 or 17 db per kilometer. That, at helium neon wavelength. And at that point, it was definitive. We knew we had — the length was sufficient that I was outside, well outside my experimental error on measurement. And the story that I’ve told, very often, was that, in fact, I can’t give you the date any longer. Somebody asked me for the date. And, I’ve stuck to my story. It was a Friday evening, late, and we had heat-treated a fiber, everybody had gone home for the weekend, and I, for some reason, wanted to test this fiber. And I put it in the rig, lined the fiber, was moving the laser spot across the core of the fiber to find the center, and when the laser beam hit the core all of a sudden there’s a bright light coming back and hitting me in the eyeball.
That never happened before. And, it took me a few minutes to realize that the light had gone all the way to the other end of this fiber, reflected, four percent of it had reflected off the other end of the fiber, and it came all the way back and was illuminating the core. I could see that the core just blazing at me. And the small laser spot in the middle. The core was actually larger than the laser spot. Hitting off my beam splitter and coming back and hitting me in the eyeball. So, I knew that something dramatic was going on about this fiber. And, looked at, I had a card opposite the end of the fiber and the laser beam was blazing out a beautiful single-mode pattern on the card. I made the measurement and ascertained that it was below the twenty db. So that was, and I can’t describe for you the euphoric feeling. That I felt that night. And, I went rushing out into the hallway. Nobody there right? Who can I tell? And, I heard the elevator door open and close. I was on the first floor. And low and behold, who should come out the door but Bill Armistead, the director of the laboratory. I’d maybe met him once. Didn’t really know him from Adam. He didn’t know me. But, he knew I was excited. And, I called him into the laboratory to see this tremendous thing. And, told him the results and so on. I’m sorry to say, I don’t think he really comprehended at all what we had. And, he made some comment to me, this was going on, and he said, “Oh that’s great, Keck. Keep up the good work. And remember that iron is a terrible impurity in glasses. And absorbed.” Something to that effect. But, in terms of the impact of the moment, I really don’t think that Bill understood. And, he was glass materials, glass chemist by training.
Lassman:Well, I’m curious though, what was management’s view of this? What was their input? Was he…?
Lassman:I assume they were aware of what you were…?
Keck:Well, they knew, I mean this was a small project. A single scientist and part-time on the part of Schulz, and part-time on the part of Zimar. I mean, Pete would, Pete was really investigating other glasses, and so, you know, once a quarter I’d go up and spend an hour shooting some soot at me. And, Zimar would spend a day doing the fiber draw. And yes, there was definitely background work that they were doing, and we were all in communication all the time. I was advising that I was learning and where I was going, and so on. But, yeah, it was still a single person and a technician project. There was not large expenditures of money going out. There, Bob had a chart that we kept plotting our progress. Once a year we had to do an annual review before Armistead, and the business leaders, they’d come up to the lab and we’d have an all-day review. And, for the most part Bob spoke for the project. I mean, in the second year I was allowed to say, speak at the review, and honored to stand before Armistead and his staff and do one of these talks. So, it was a big deal. This was the epitome of research, if you will. You didn’t go outside and present your paper. You presented your paper before senior leaders of the laboratory. But Bob had this chart where you charted the progress and the reduction of attenuation with time. And, we were making some progress but there were constantly rumors floating around that we might have to give up the project or start doing some other things, hints that maybe they’d spent enough money on this in pursuit of this thing.
Lassman:Rumors from where?
From Bob, or from other scientists, and so on. And, at that time, I was naïve as to how projects were selected, and didn’t know the operation of the laboratory at all. Later, my colleague in manufacturing, Charles “Skip” Deneka became the chief technology officer. Skip Deneka was, had somehow found a list of, a project hierarchy list of projects in the laboratory, circa this time frame, and out of twenty projects the waveguide project was number seventeen. So, it was very close to the cut list. About this time, the catalytic converter activity at Corning was reaching fever pitch. The automobile companies had been told that they had to reduce the pollutants, and so on, and I can recall before we got our low-loss fiber, Bob called the department meeting and basically said that half of the department was going to have start working on catalytic converters. What ideas did anybody have about how we do this? And, happily, he allowed me to continue with the waveguide project, and others in the department began trying to do some work on the catalytic converter project. But, we literally moved half of the laboratory work on catalytic converter. That was just how I later found, Armistead wanted to do things. You know, underwhelm it. And, of course, we built a factory in anticipation of the market. We didn’t have a process and didn’t have a material. Age-old story at Corning. They, every time, come up with the solution.
They just get all the minds, smart minds you can, working on the thing, and focus, and somebody comes up with it. And, indeed they found the corrugated material and they found the processing way of proper dye that would allow the extrusion to take place, and how you did it, and came up with the process, got in the plant, and came close to shutting down Ford Motor Company’s production line. But, we made it. And, the gentlemen are in Hall of Fame that came up that process. But that, that was the climate, circa the time that —, just before we came up with the breakthrough, masses of scientists in the lab were moving over to another project. A big, big focus on that.
Lassman:That’s the development of Circor, is that right? Or Cel—, materials?
Keck:Circor was a…
Lassman:I get those confused some.
Keck:A glass, gas turbine.
Keck:Material that we came up with, a heat exchanger…
Lassman:Actually, getting the details for this is ideal. But, two, at least two, well actually now three questions come out of this. But, the first one is, at the time that you’re doing this research is there any, are you participating outside?
Lassman:What’s going on in the larger community? You’re clearly on a focused path here.
Keck:So Corning wasn’t in the habit of sending a lot of scientists to technology meetings. But, they began to break that. I was interested in going to conferences, and keeping up with events in the technology. I remember going to a CLEA meeting, it was a forerunner of the CLEO conference that’s still going today. Optical Society, IEEE LEOS, conference on laser engineering and… What’s the “O” stand for? I don’t remember anymore.
Keck:Maybe. Could be. I guess. Suddenly it escapes me. Well, anyway, it was CLEA back in those days. The Conference on Laser Engineering Applications. Largely a laser conference, but you began to piece together the fact that other laboratories were investigating this transport problem. And, then there was an SPIE meeting that I attended where…
Keck:Society of Photometric Illumination Engineers. It’s a sister society to OSA. But, it’s an entity unto itself. It’s its own organization. And, it sprang up as an engineering counterpart to the more academic OSA. And that conference was where I heard about these military applications of fiber optic illumination bundles and things like that. You began to piece together the fact that other laboratories were working on it. There was another, I think, a Ceramic Society meeting that Nippon Sheet Glass’s, Dr. Ichero Kitano gave a paper for when Nippon Sheet Glass was partnering with NEC. I’m not sure. Something around, I guess maybe it was NEC. I was going to say MTT as well. But, maybe it was just the two of them. But at any rate, he gave a paper on the Japanese fiber work. And they were working on a fiber that later became known as Selfoc, and it was a gradient index core. They were using a double crucible approach to make the fiber, where you literally melt glass in two concentric crucibles and allow it to flow together. That’s the other way of making fiber that was around in this day and age.
Lassman:As opposed to…?
Keck:As opposed to the rod.
Keck:So, Kitano gave his talk and I think I remember that he was at 500 db per kilometer. That was the loss rate for his fiber. And that was one of the first measures that we had of what other people were doing.
Lassman:Is anybody else using fused silica? You’re the only one?
Nope. Nope. This is the contrarian approach to business I think it was at that conference that we learned that Dave Pearson, David Pearson of Bell Labs, had a twenty-person group, about that, working on melting glass at Bell Labs, trying to make low-loss fiber. And they had reported some of their work. Their approach was to buy purer and purer chemicals from the JT Baker. I remember a representative from Baker kept coming to some of these early conferences talking about how many nines purity they had on their materials that were used in glasses. And later we found that George Newns at British Telecom Labs was also trying some double crucible work. He was a glass chemist at British Telecom. So, we had learned that Japan, Inc., and later MIT, we found had a sizable effort. Bell Labs had a large group. Newns had a fairly substantial group at British Telecom.
So, those were the competitors. The telecom companies were obviously trying to come up with the next generation telecom. That’s in their charter of course. But, all of them were taking what I call the engineering approach. They were taking the very best that anybody knew about glass, at that instant in time, and trying to improve on it, to the next level. The purest glasses that people were melting were for camera lenses, high technology optics. And these were typically multi-component glasses. Had multiple chemical constituents. Typically, calcium, silicates, and things like that. Sodium was an added that’s often used, maybe a little phosphorous. The glass industry learned about finding agents over the years. And so, you’d put a pinch of this and a pinch of that in and you get better melting properties. So, the thrust of all of these groups was to use a distinct process, largely double crucible, and — well now, Kitano, Kitano at that time was , but Newns was approaching a double crucible, and I’m pretty sure Bell Labs was double crucible. But, they were trying to purify these four and five component materials that you’d put in the melting batch.
Every one of them had to be seven nines, or nine nines pure. I mean, parts per billion, we later ascertained were required as pure, allowed impurity level of the irons and the coppers and so on, in order to get these loss values. So, here they were trying to, as I say, just increment their way forward. And Bob, the thing I always really learned from Bob was this contrarian notion. If you do the same thing everybody else is doing, the best you can do is tie. And he said, you know, we’re going to go in this direction. Pick the silica from, as I indicated earlier, his research on scattering. I knew that this was, A) a simple material, just two components, and B) we had the vapor deposition process to make it, and later found that it had this distillation capability, that made it a very pure process, pure material. And, the second contrarian notion was this adding impurity, of all things, to make your core. And the third was to depart from the normal ways of processing the stuff and go to this thin-film process to get high quality interface and the right core cloud ratio. So, those are the three contrarian, swimming-upstream, things that nobody was trying, and we were the only ones doing it.
Lassman:Did any of these competitors try to pick up on what you were doing?
Keck:Well, they didn’t know. But there are different levels of classification. Have you heard this story? In military circles, it’ll appear in the New York Times two weeks after its classified. But, a materials company, if they stamp something proprietary, it’ll never see the light of day. That was the mindset. So, one of the things I learned coming out of academic work is the level of proprietariness, that you kept things. We gave no papers. We told nobody about it. We went and we listened to these conferences. Just listened. And people would come up to us, knowing we were from Corning, and asking us what we were doing. And, “Oh, just curious.”
Lassman:I guess another related issue is to the extent to which some of these other competitors might have seen Maurer’s paper? Was that something they might have passed over?
Keck:Oh, it was so long ago. It was ’53, after all. Mallitson was the guy at NIST that did the silica. Just thought of his name.
Keck:I. H. Mallitson.
Lassman:You got the low-loss below twenty decibels?
Keck:Yes. Got it once.
Lassman:I want you to take me through then how the fiber optics moves up from number seventeen on the list and begins to move up on the list. And there are a couple of things that come to mind. One, the general excitement about fiber optics in the 1970s, and how management’s view changes? You mentioned earlier that Armistead apparently didn’t really recognize what you had accomplished; how does that then become a priority?
Keck:You might want to soften it, when you write this. Was uncertain. By this time one of the background things that had happened, and I can’t give you the date. I may have, well I’m sure I have it in my lab book someplace. We had struck up a relationship with Bell Laboratories. Bob had — I don’t know how the interaction came about, but somehow Stu Miller at Bell Labs was running the Crawford Hill Laboratory, and that was the laboratory that was doing all the optical, well laser work, and possible optical transmission work. And, they were working on…
Lassman:Should I pause this for… [Keck referring to images on his computer]?
Keck:No, that’s fine. Keep it rolling. I’m showing a slide of transporting laser beams through pipes. Here, they couldn’t go through the atmosphere, so they had the project in their microwave laboratory was what this adjunct laboratory was about, and they were, under Miller’s leadership, they were moving into the optics arena. So, they took these old microwave pipes and put lenses periodically in them to refocus the laser beam and keep it guided through the pipe, and mirrors at the corners, and so on. And so they, the fellow that came to be a colleague, Detlef Gloge, was looking at the lens guides. Peter Kaiser was looking at a gas lens guide where they’d fill the pipes with, I guess rare earth gas, and by putting heating coils at periodic intervals along the pipe, the gas in those regions would be less dense out near the periphery. Less dense corresponds to a lower refractive index. So, gas in that region would have high refractive index near the center of the pipe, lower refractive index near the outside, and would form a lens, a thermal lens.
Lassman:Well, that’s interesting.
Keck:So it refocused the laser beam. So, Peter Kaiser was working on lowering that thermal lens thing, and Detlef Gloge was putting physical lenses in the pipe. And, I remember on one of the first visits we had where we began to work together, Detlef told me about the, what a sensitive seismometer he had. He could pick up a truck on the New Jersey Turnpike ten kilometers away from the vibration of the lenses in his pipe. So, he was not all that enthused I guess about the prospect of it. Those are words, those are my words, not his. So, this laboratory at Bell Labs had a substantial effort on trying to figure out how they were going to transport laser beams over long distances. Pearson was in the materials group at Murray Hill trying to come up with fibers that would do the same thing, glass materials. Well, circa ’69, Corning needed the transistor patents for the venture, the Signetics venture, and we signed a cross-license agreement with Bell Labs. And, one of the pieces of technology that we put into the cross-license arena was anything that came out of fiber optics.
Lassman:Hmm. I thought that transistor licenses could just be purchased.
Keck:The powers that be, whoever negotiated it, set it up with a cross-license. Maybe there was money that transferred. I don’t know. But, future technology was in the mix as well.
Lassman:This is before you get the low-loss?
Keck:This is before I get the low-loss.
Keck:Bob somehow associated with that, I assume, had met Stu Miller. There was probably an entourage that went down and negotiated it. Whether Bob was part of that, or whether Chuck Wakeman was, I suspect Chuck perhaps, it was a pretty high-level negotiation. But, somehow or another Bob met with Stu Miller and we began to have periodic meetings with a group at Bell Labs — Henry Marcatili, Detlef Gloge, Peter Kaiser, Art Tines, Charlie Burroughs, Tina Lee, and so those were, there were others whose names escape me. Stu would get his group together. They would report on some of the stuff they were doing in fiber optics. These were all under nondisclosure agreements of course. And, we would talk about what we were doing in this area.
Lassman:Did you give them the details of what you…?
Keck:Bob probably talked about our scattering work from the, you know, early times. I think I gave a talk on measurements, perhaps. Felix Kapron was in our group. Felix was a theorist. And, so, I’m pretty sure he came down to talk with Marcatili. He was a master theorist. Schulz never came down. He was not part of the, part of the team. Didn’t want to have any notion that silica is the direction we were heading. If Schulz came down, he was a young researcher, and the people in the glass industry probably knew that he was hired to research high silica glasses. So, he never came down.
Lassman:So you weren’t showing anything about the process, or anything like that?
Keck:No. We did not. And a lot of it was, you know, just the early meetings of getting to know one another. So, there wasn’t a great deal of substantive stuff that transpired. But, the relationship had been set up. When we got the low-loss fiber, one of the first things that we did, Bob took that fiber down to Bell Labs. Bob was, he’s a German background, and never got very emotional. And, I remember we were doing a draw over in the development building. He, I’ve forgotten how he got the fiber down to Bell Labs, but some of the questions that came up around this first fiber was 200 meters of fiber that we had was, “Well, Don, are you sure your measurement’s right?” “I mean, this is dramatic. Could you be mistaken?” So, he took it down to Bell Labs and they corroborated our measurement. And, I remember the day that Bob must have been talking to them, and I was over in the development building sweating away. This is mid-1970 in the summertime. And gosh, that development building would get hot.
Keck:This furnace is radiating and it’s up at 2000 degrees C as we’re drawing fiber at that temperature. I had to melt the silica, of course, and Bob darn near embraced me. He came over and said, “They corroborated our measurement.” I’ve never seen him quite as emotional, and demonstrably emotional as that time. So, that was the first corroboration. And, Art Tines and I then — well, let’s see. No. That’s not quite the sequence. Oh boy. Now I’m, now — we published our work, as the world knows. And, I don’t know whether the publication came out subsequent to Bell Labs duplicating our measurement. I suspect, I suspect Bell Labs had corroborated our measurement before we published the Applied Physics Letters paper. But, we sent in a paper to Applied Physics Letters.
Lassman:I have your list.
Keck:In reference to ’70 or ’71. It was in the late —
Lassman:It says Kapron, Keck, Maurer, Radiation Glasses and Glass Optical Waveguides. Applied Physics Letters, 1970.
Keck:Yes. So, the fall of ’70 we wrote the paper. Notice that Schulz’s name is not there.
Lassman:Yes. I do.
We said nothing about the material. Kapron was off to the side just doing theory. He had looked at Marcatili’s paper on bending losses. And I don’t think it was fiber waveguides but in planer waveguides. Marcatili at Bell Labs had written a paper, and Felix was following everything that Marcatili did, and it was written up in BSTJ [Bell System Technical Journal], probably. So, Felix wrote a theory of the bending loss in round fibers for us, you know, the scale of Marcatili’s results or something. And the way we publicized the result. Well, two-fold. One, Bob went to a conference in England. And two, we had the Applied Physics Letters paper. And, it was a bending loss paper. I basically looked at the loss increase as you bent the fiber in tighter and tighter radius, and took the experimental data and Felix got the shape of the curve for us and we simply scaled it to our results. And, it followed the results beautifully. And in we, the passing comment in the text was, “Oh by the way, the fiber that we used in this work had a loss of less than twenty decibels per kilometer.” That’s the first announcement that the world had. That this fiber existed. And Bob, we’ll have to look up the dates. Bob gave a paper, it was a conference on microwave telecommunication, in London. IEEE conference on microwave telecommunications. But, there was a section on optical telecommunications.
Most of the conferences on microwaves. And, at the time, interesting history again — engineering approach to things - long distance telecommunication was taking place over microwaves, and beaming microwaves from tower to tower. But, there was work going on higher transmission cables. And, it’s well known in the electrical engineering that the larger you make the cable the lower the losses are as the frequency increases. The frequency dependence is the square root of the diameter. That’s not right. I take it back. Loss falls off, is the square root of the frequency. So, as you go to higher frequency — I’m sorry, loss increases as the square root of frequency. So, as you go to higher frequency, why your losses are going up. And the rest of the theory says that if you make it bigger, why you can mitigate those losses. So, the whole cable industry, wire cable industry was making, was investigating the possibility of two-inch diameter helical microwave guides. And, one of the things, one of the reasons Marcatili had written his bending paper was that you find that the bending losses in these things would be, you know, enormous.
The bend radius you’d need would be circa that on a superhighway. You could follow along the highway system and the losses would be sufficient, the bending losses would be sufficiently small. So, the cable industry is experimenting with how does one make these cables… Bridging glass pipe and somebody came up with the idea, well, we had a microwave laboratory down in Raleigh. Looking at low-expansion ceramic materials as microwave cavities. Gerhard Megla ran the laboratory and he was a microwave engineer. So, we had a whole group down there. Well, at any rate, the notion was to take our glass pipe and coat it with a metal on the inside and make a helical microwave guide out of glass pipes. That didn’t ever really get started, but the rest of the cable industry was looking at these big two-inch diameter coaxial cables, and stringing them all across the countryside. And, I attended a conference, it may have been that CLEA conference I mentioned, in 1969, down in Washington, where the major telecommunication companies in the world, NTT, British Telecom, AT&T gave papers on their engineering tests, field tests, of these two-inch diameter microwave guides. Transmit 300-gigahertz carrier in the sub-carrier multiplex of sixty channels. I don’t know why I remember [Laugh] all these things.
And, each one of them, you know, they were about the same stage of deployment. They were actually had these engineering tests going on in the field. And, I’m sure they talked about it at the conference that Bob went to. And, Bob gave our bending loss paper at this microwave conference. Oh, by the way, twenty decibels per kilometer. And, apparently, the conference was just stunned into silence at hearing this. We began hearing speculation on the part of these helical guys that now at least we can fill the pipes with fibers. So tremendous distance to go from where we were when we reported these results to actually getting a business.
Lassman:It’s clear though that in relation to the coaxial cable business that that’s not going.
Keck:Well, I don’t think one can even say that. But, certainly, people went back and began to rethink this engineering approach using better cables. I had talked to a fellow that was at British Telecom, in the microwave division, some years later. He got, or was in management, and we later had a, we bought a part in British Telecom Labs is what we did, during the dot com [boom]. And, I met this fellow and he recalls that they literally shut their program down after our result and put the effort on the optical approach, and beefed up Newns, George Newns glass department, and pushed on that.
Lassman:I just have a question about the technology, to make sure I’m clear. You mentioned, in the microwave field, you were transmitting from tower to tower? So, it’s not clear to me where the cable comes in?
Keck:Well, they were going to supplant these beamed things with those pipes.
Lassman:With the pipes instead? Okay. I see.
Keck:Or, a different part of the network, or something like that.
Keck:But that was how they were going to solve the problem. The ability to carry telephone traffic was, you know, the wires were simply running out of steam. You couldn’t go — chart in that presentation. I can get you a copy of that if you want it. Well, so now we’re —
Keck:Where do you want to go? Because I can keep…
Lassman:Well, let’s, you mentioned the new cross-license agreement with Bell Labs? So, how do you move from ramping it up to production and capturing a market? And, where does management get on board?
It took twelve years. And, again, dates are going to evade me. Chuck Wakeman took a new job. Les Gunderson came up from Raleigh to run physical research. He was a young up-and-comer, microwave specialist. I remember one of the first meetings that I had with Les, we had both attended the first integrated optics conference that the Optical Society put on out in Las Vegas. And, P. K. Tien, at the start of his paper said that — Las Vegas was a perfect venue for this conference, because anybody betting on integrated optics in 1971 was a real gambler. Or something to that effect. And, Stu Miller — our patent had issued. That was it. In Europe, Danish patents were published six months after receipt. And we had filed broadly inside the tube process and the sequence was that Bob, Bob and Pete Schulz filed the patent on the material. Pete and I filed the patent on the inside of the tube process. Those were published, and Stu Miller had seen them. Because at this point Bell Labs doesn’t know in any of our interactions how were making them, nor were we — it might have been barred from our discussion. We weren’t to talk about materials. I’ve forgotten whether that was a stipulation.
It was primarily associated with measurements, and theory of waveguides. I think there was probably something along that line. Well, at any rate, Gunderson went with me to this integrated optics conference, and Stu Miller came up to us an asked if we would give a post deadline paper on the process for making the fiber. And, Les had to make the decision, and he was just brand new in the job. And, I didn’t know any better. We didn’t. But, I do remember Stu pleading with us to give this paper, to get the word out, and the technology. Subsequent to that, one of the next things we did was Art Tines and I — well, we made our measurements at the helium neon wavelength. And, Art Tines might have had a spectral response measurement. I don’t remember.
They certainly did subsequently, but whether they measured the spectral response of our fiber I don’t know. But, the notion came about that we were going to co-publish with Bell Labs on the spectral response of the fibers. I had my quickly assembled spectrometer to measure the spectra response of the fibers. And, the light, well to get a white light source to do your spectra response, and get enough light in the single-mode core, I mean there just weren’t light sources that you could do that. A laser would be able, you know, you could focus it down, to the diffraction-limited spot. But so we, I had learned about zirconium arc lamps back in Michigan State. That Clarence Hawes, or Duane Hawes had introduced me to those and told me that they were the brightest, optically brightest source. So, I got one of these zirconium lamps and couldn’t use a grading spectrometer. It would disperse the light too much. I eventually found the small prism spectrometer they had laying around the lab. And that had sufficiently low dispersion but would still give you enough of a look at the spectral response. So, I measured the spectral response, and Art Tines measured the spectral response, and we published a paper to Applied Optics, on the Spectra Response of Low-Loss Fiber. That was, the first co-publication that we put together. This, I think hit in ’71, maybe. Is there an Applied Optics paper there, someplace?
Lassman:Yes. Kapron and Keck, Post-Transmission Through a …?
Keck:No. Should be — I think it was before that. Maybe.
Lassman:Minor Fusions and Dispersion Effects on…?
Keck:Oh, here it is.
Keck:Seventy-two? Okay. Later than I was remembering. So, when did Felix and I do the thing on pulse transmission? Seventy-one? Okay. Then, okay, October of ’71. Yes, we had low-loss but this fiber, remember, was heat-treated, fragile as all get out. We’d already broken it once. I wonder where that fiber is? Did I give that to, I think the Corning museum of glass has that original fiber. (And, it might be on display now. But, over the years it broke, and broke, and broke, and there were ends sticking up all over — we wound it on a paint can, one-gallon paint can. Oh gosh, Tom, I’m going — you’re going to be here for a week. Just tell me when I’m getting too far afield. So, we knew we had a fragility problem. We knew we’d have to duplicate these results. We tried several times. And, I think the next two attempts failed. By this time, we’ve filed the patents. Schulz-Maurer on the material and Schulz-Keck on the process. In that first paper, of Maurer and Schulz publication, we listed the materials that Pete knew he could deposit in the fiber. Germanium was not one of those, because when he tried to do Germanium by flame hydrolysis, and by that, I mean not a soot process, but actually laying down a clear glass to begin with, he was not able to keep the germanium from vaporizing. But, somewhere along in that ’71 or ’72 time frame we tried a couple more attempts to reproduce it with titanium. We couldn’t. I don’t know how or when Pete realized that he could retain some of the germanium in the soot process, but he did. And so we made a multi-mode fiber, I think ’72. Isn’t there a paper there on “The Ultimate Low-Loss of Fiber?”
Keck:Was that ’72?
Lassman:Applied Physics Letters, on the “Ultimate Lower Limit of Attenuation of Glass Optical Waveguides?”
Keck:Yes. What’s the year?
Keck:Three? Okay. So, somewhere in that ’72, ’73, timeframe we moved over — well yes. Okay. That’s the sequence. I wish I could tell you the date in that ’71, ’72 time frame. Probably, after the integrated optics conference which was in January of ’71. So, sometime, I’m going to say ’71 at this point, a contingent from Bell Labs came up to Corning. I don’t remember whether Chuck Lucy and Bob invited them up to try and excite Corning management about the project, but I suspect that might well have been the case. But, at any rate, Stu Miller, Alan Chenoweth, Bill — oh gosh. I don’t recall that name. Another fellow from Murray Hill, that ran the Murray Hill Lab. Chenoweth ran Holmdel, maybe? — Or some large group at Holmdel. And Bill, whose name escapes me, came out of a materials staff at Murray Hill. And then Miller ran the small adjunct lab at Crawford Hill. And, the three of them came up.
Lassman:Would it have been Fisk, from Murray Hill?
It might have been. Might have been. It could have, could have been. At any rate, they came up and I want to say Armistead and — it was a high-level meeting. I suspect Chuck Lucy probably orchestrated it to basically talk to Corning management to indicate how important this was, and therefore that Corning management should apply more resources to moving this thing forward. And, I don’t remember whether Armistead was in the meeting or whether it was the directors, Chuck Wakeman. Well, if it was Chuck Wakeman, yeah it had to have been before ’71 then. Well, maybe it was, maybe it was late ’70. Well, at any rate, that meeting took place, and I wasn’t part of that meeting, and so I don’t know what went on, or what was said or anything like that, but nevertheless we began to add more people to the project. The development group, Zimar’s fiber drawing group was enhanced too, and Pete was allowed to add some people. I think Miles Vance came back to work on the thing in our group. At any rate, the effort began to increase. But, we still had the problem of, why this thing was fragile. By this time, well the Tines paper we found that water was absorbing strongly at 1,300 nanometers. Yes, 1,370 nanometers. And that ultimate lower-loss paper, my spectroscopy came back into vogue, because what I found was that in these long path lengths we were looking at the harmonic bands, the overtone bands of the fundamental.
Water, the stretching vibration takes place at 2.7 microns in infrared. That’s the fundamental. The second harmonic occurred at twice the frequency, on a shorter wavelength. And typically, it’s an order of magnitude lower absorption coefficient. And the third overtone would be ten times lower than that. And so on. So those bands are, you know, fall off very rapidly. The selection rules of probability of excitation goes down as you go up to these higher orders of coefficients, or in the higher returns. Well, the 1.7-micron band is the first overtone of water. So, it’s down an order of magnitude from the fundamental. And yeah, we were seeing it very nicely. And, low and behold, there was a band at 950, which was identified as the third overtone. The gallium arsenide lasers emitted 900. And so here’s this big bump in that spectral response. And, we’ve got a water problem that I identified. And there are some combination bands of the 0H has a perturbation from the silicon matrix vibration, so there’s a 2.4 micron band and I could see that appear over time.
On the first response, I don’t think we got out far enough to see it. I’m sure we didn’t. You know, but eventually I got spectral response all the way out to, you know, the fundamental. And beyond. So, we had the fragility problem. We had the water absorption. We hadn’t duplicated the titanium. And, let’s see. We said ’73 is when we did the ultimate loss, so it must have been sometime in ’72 that we made the fiber with germanium in the core. And, by this time Bell Labs had interaction there with some of their semiconductor laser guys had indicated that the active-A in the semiconductor laser was going to be fairly large, fifty, sixty, seventy microns long, emitting anywhere in that region, and if they were going to align it to the core, sometimes it would move across the junction and you wouldn’t always emit from the same spot. And so, we said “Couldn’t you go to a multi-mode fiber?” So, that ultimate loss paper was our first attempt at a multi-mode fiber, large core thinner cladding, but still enough to protect the wave, and we added germanium as it opened. And that was such a wonderful fiber because there was no reduced state of germanium oxide. It went in a perfect stoichiometry into the matrix. And so, we didn’t have to heat treat it.
Lassman:Unlike the titanium-ox…?
Keck:Unlike the titanium.
Keck:So, in ’72 I traveled to Geneva, Switzerland to give a paper, the first European optical communication conference. And we admitted that titanium was our dopant. Germanium was now our new dopant. Didn’t tell the world that.
Lassman:The titanium for the single mode?
Keck:We were using — well germanium was going to be the dopant for both of them.
Lassman:Okay. I see.
Keck:But, because of the Bell Labs input on the sources we shifted our total thinking away from single mode, and began making multi-mode fibers now.
Keck:So, we gave papers, and got out there, and so we would, you know, could talk to people. We were being viewed as responsible scientists, in leadership positions, and so on, and so on, but we were always giving data on one generation before where we were actually working. Didn’t I fail to point out the Applied Physics paper, and Kapron’s name appears, Maurer’s name appears, my name appears, Schulz doesn’t appear. That was conscious. We knew that he might be known in the glass circles, he was working on silica, and we didn’t want to give away the possibility. And we’ve taken a lot of flak on that subsequently.
Lassman:How did he feel about that?
Keck:Burned. And eventually it hurt him. But later I found out that all is not as serene in industrial research as it might seem. You hear about the political issues in academe, and by in large, Corning was a chemical company, and chemical processing, glass, a materials company, and the physics group was kind of viewed as an important discipline, but not in the mainstream. So, lots of papers were coming out on all sorts of glass stuff. And, the physics group was generally not included in publications that came out of the glass department. And so on. So, it was a little, little bit of friction between the departments that occurred or what existed at high levels. Pete and I were totally naive young scientists. All we were interested in was, you know, getting the word out, publishing papers, and doing the work, and so on. So, initially it didn’t bother Pete. But, later on it clearly had bothered him, and eventually we were able to rectify it and use the co-inventor on the optical fiber in the National Inventor Hall of Fame, and so on. But it was a conscious, conscious choice.
Keck:I meant to mention one other thing on that paper. We got the review back on the Applied Physics paper and the reviewer, I don’t know who it was, but the reviewer rejected the paper for being published in Applied Physics Letters. And his comment back to us was, “It is really difficult to visualize.”
Lassman:“An amorphous solid with scattering losses alone below twenty as per kilometer, much less the total attenuation.” No.
Lassman:He just didn’t believe it?
Keck:Didn’t believe it. Didn’t believe it! That was one reviewer. The other one was okay with it, and the editor happily let it go through. I still have no idea, but I delight in showing that at conferences.
Lassman:That’s — convincing the …
Keck:Our peer review process is something.
Lassman:Convincing others of what you’ve done in the lab can be tough to do. Well, you’ve taken me through the early development, and I would just like to shift the conversation a little bit, but I wanted to get just a little bit more information about the market, and what the conception is. You’re clearly trying to now convert it from a laboratory discovery into something you can manufacture and produce. I just want to get a sense of what the other side is and what the view of the market is, and how that develops.
Well, I’m not sure … — Chuck Lucy was, in many respects a genius. Not only was he a top-notch physicist, but he understood people, and so on. And, he was clearly the business development director for this thing as it moved forward. So he and Bob sort of, Bob was the technical, and Chuck was the marketing person. They started adding, probably about the ’73 timeframe, started adding marketing help with Chuck. He got an assistant that helped him. When did I do the SPRE paper? Seventy-five. So, Chuck was already well on board by then. He must have, he must have relinquished his new development position, new business development position in the telecom group, and pulled out to spend total time building this new business. And, this, Chuck, I never saw a market projection from Chuck. Later on people were speculating that Chuck knew that if he put any numbers out they were always going to be scrutinized, and so on, so he just never put a number out. He said, “The market’s big. The trunk is one, the volume of the metropolitan area will be ten times larger than that, and if you go into the local loop, it’s ten times larger than that.” And, “Think of the amount of copper wire that you, for telecom, or for telephones year by year, and we’re going to replace that,” and so on.
So, I don’t recall ever seeing any market numbers for that. And, probably the company as a whole wasn’t as disciplined as they would be today in investigating. I don’t know. But, to Chuck’s credit, in this general timeframe, armed with the papers that, you know, we had given — and by the way there’s another important paper in there that I want to come back to. (Chuck had gone around to five major telecommunication firms, Furukawa in Japan, joined with NTT. No. With Fujitsu, as the two industrialists, and NTT as kind of the over-the-shoulder observer. Siemens, in Germany, Deutsch Telecom, Plessey in Great Britain, British Telecom looking over their shoulder. CGE, Company Generale Electrique in France, with the CNET looking over their shoulder. And Pirelli in Italy. With Sault over their shoulder. And had arranged a deal, a joint development agreement, and I don’t remember now exactly when those were signed, but they were pivotally important in the whole evolution of the business. We’re starting to spend quite a bit of money.
There are probably ten, twelve people on the project. And, for a company the size of Corning that was enough. You don’t spend that sort of money for a long period of time without knowing something about the market. But, Chuck very cleverly had done joint development agreements with each of these companies, industrial companies, and the gist of it was, you know, you’ve heard what we’ve, what we think we have, why don’t you pay us $100,000 a year and we will consult with you and tell you about the progress, and glassmaking, and teach you about glass, and what we know about it, and you teach us what we need to know about telecommunications. And how the product life changes and so on. And he signed these agreements with all five of those. We had a half a million dollars coming in, which was frankly big money for a project this size. And [for] that he didn’t have to show market numbers. He had these signatures on the dotted line from senior people in each of these companies. And in fact, it was a five-year agreement.
So, we had to do this for the next five years. At any time they could take the option of forming a joint venture with us to produce cables, I think, was probably how it was structured. Siemen’s jumped at the chance immediately, and we formed Siecor Chemical. And the rest of them had the option to do it later on and Plessey and BICC was the other couple. There were two in England. Plessey and British Calendar Cabling Company. And later we formed a joint venture, Optical Fibers, Incorporated, in Great Britain. Pirelli took a license and just went [on] their own. They didn’t do a joint venture with us. Paid us royalties on the patents. Furukawa did the same after a bit. But we, it was a heady experience for a young scientist. You know, you’re now four or five years into the mission and I got to travel all around the world. The regimen was that every six months we’d have a meeting. One of them may come to us, one by one. And the others we’d go to their place and tour around.
So, I toured Europe, and we’d get our presentations ready and go out and teach them. I met some wonderful scientists in all these companies, and we collaborated with a number of them. There’s a paper by one of the French scientists that I co-authored. And, others. And, they began to talk to us about the real issues that we were going to face. “How are you going to connect these things? How are you going to get the light in and out?” “This stuff breaks. Glass breaks. How do you, put a telephone system together with something that breaks?” So, we immediately began to do, Bob launched a program, and brought in some other folks to worry about the strength of the fiber. After we got rid of the germanium; we had nice pristine fibers coming out. We knew we’d have to coat it, so one of the cable companies counseled us on what materials to put on. So, we started putting lacquer on the fiber, a thin layer of lacquer. And, you could never get it on thick enough, and you always had little gaps in the coating process, and so on. But, we were starting from scratch and learning how to cable a fiber.
And, we knew from — oh no. [sounds like “kina”] was the first one. I have no idea what that is. “How fast can you draw and what curing time?” We began setting up a development group whose sole purpose was to look at draw technology. And we had another development group that was aimed at how do you deposit the stuff and get better uniformity on putting the stuff down? By this time, by the way, we had invented the outside deposition, vapor deposition process. Pete and I had realized that trying to put this stuff inside a tube was just not going to get the length of fiber that you were going to want. So, we simply inverted the process, and said, “We’ll start with a rod one and pull it out. And, we’ll deposit our core, you know, a few layers of core material and start laying down silica with these burners.” And that began to catch on. The development group started trying to figure out how to make bigger and bigger pre-forms. The first, we’d get a kilometer out of these little six-inch inside tube blanks. Interestingly, Bell Labs stuck with the inside process. So, John Machesney, you may have heard that name in technology, John claims to have a variant on the theme.
Instead of having his flame inside — well external to the thing and shooting in, or has a plasma inside, he said, “I’m going to heat the tube on the outside and simply do the chemistry inside the tube.” And he passed his materials through the inside of the tube and heated it on the outside, called it a modified chemical vapor deposition. Because in Pete’s and my patent we had said, you can put this material on the inside of the tube by any of a number of techniques, one of which was chemical vapor deposition. And, flame hydrolysis was one, and radio frequency sputtering was one. And any that we thought would lay down a thin film we listed in the patent. Well, Machesney got a patent. We never litigated because there’s a cross license. It didn’t matter whether he had a patent. Machesney has gotten, in my opinion, far far more credit than probably is warranted based on the work that he did. But eventually, he duplicated our result, with this inside process. And, they could get two kilometers out of the tubes. You know, we got one out of the little short thing. So, we knew we had to do something. The outside process we eventually scaled it up so we’d get ten kilometers. And, I can’t tell you how many kilometers they get off a blank today. So, those things were going on in the background.
Lassman:You, well you had mentioned, you know you’re talking about Bell Labs now —
Lassman:— and I know this is a story that’s told in the Corning histories —
Lassman:— that came out, but you know, you’re sharing the technology with them, but it’s a telephone monopoly -–
Lassman:— so they actually, they control the entire telephone system.
Keck:Yes. And Western Electric is making two billion kilometers of wire a year.
Lassman:Yeah. That’s their captured producer. And then the breakthrough comes in, with the breakup of the Bell System in the early ‘80s, that opens the door?
Yeah, but that whole time period of ’75 through ’81, ’82 when the break up occurred, Chuck Lucy is dancing like you wouldn’t believe inside Corning, because all the nay-sayers were saying, “How much of Bell’s business are you going to get?” You’re going to load level their production, Western Electric’s production, and when they’re out of capacity, they’ll give you a little volume, and when they’re in capacity, why they’ll shut you off. And so you’re just skimming. And Chuck had to really bob and weave a lot on the nay sayers. Happily, the chairman of the corporation, Amo Houghton said, “I don’t care. I want to be in this business. Oh by the way we’re going to build a plant in anticipation of the market.” So, we set up a pilot plant in ’75, down in Irwin. We wanted it close to the laboratory so we could transfer the technologies. And as a scientist, I was responsible for the measurement QC systems that were going in down there. So, we were, you know, frequently in the plant getting it up and running.
We were making at the pilot plant, a few hundred kilometers a year. But, it was enough to send fibers to our joint venture partners for them to play with and try and cable, and do experiments on the transmission properties and so on. And, Bob was investigating the strength of fiber. And so we got a lot of papers on —, that eventually showed that even if you stressed the fiber and held it under tension, if you coated it immediately as it came out of the draw as a pristine surface of glass, that it could last for thirty years if you didn’t exceed more than such and such a tension on the fiber. And we proof tested it. We built the first cable under contract to a Naval Electronics Command. They funded us to make a cable.
We made the first six-fiber cable and used Kevlar strands on the diameter, the diametrically opposed fiber in the middle. And so, it would only bend in one direction. But, had to work in coating the stuff and cabling the stuff immediately ran into the microbending problem. If you put, never could prove that it was actual physical bending. I, to this day, believe that it was probably a stress optic effect that if the fiber passed over a nodule or a spot in the organic materials that we were extruding on the fiber and you then pulled on it it would press on the bump, or impurity, and cause a stress optic change of the refractive index and scatter light. If you had enough of those per kilometer, you’d raise the losses well above acceptable limits. I already knew about microbending. There’s a paper in here early on where I’d wound a fiber too tightly on a drum, came in over the weekend, and wanted, as early in the game, I wanted to show Chuck Wakeman.
He hadn’t seen the low-loss fiber. At eight o’clock in the morning, I called him up. Bob was on vacation. I called him and said, “Do you want to see something? Come on down.” And then I went in and turned on the laser. Well, the  had come through the fiber. It had been piping light beautifully on Friday. It turned out over the weekend they shut down the air conditioner in the summer. This cardboard drum had soaked up water and just expanded and the microbending shut the thing off. And then I started investigating the microbending, and others did too. And, we wrote a paper on the problem. I wanted to point out an important paper in here that I claimed should take great pride in publishing. This one that Kapron and I authored on —, it’s the bandwidth effects. We were the first to, and I bet it’s in Applied Optics. Is that the one? It’s core electronics.
Keck:I guess it must be.
Lassman:You were pointing out this paper, Kapron and Keck, “Post Transmission Through a Dielectric Optical Waveguide,” Applied Optics, 1971.
Yes. So, as soon as we had the low-loss and we began to assemble more people working on this, as I’ve said, other questions came up. The strength of the fiber I’ve already talked about. But, one of the next things was, well, okay, what bit rate can you send through the fiber? I knew that it was going to have to be a pulse transmission, a digital transmission of some sort. And, so, Felix began to try and look at the propagation equations for the dielectric waveguide. And, an old model of the light source and my physical intuition I think played a pretty important role in that paper. Felix knew the math, all right, but the physical intuition of what terms needed to be put into the equation was very important. But, basically, you have the dispersion characteristics of the cladding. You have the dispersion characteristics of the core. And they’re slightly different than one another. And then the waveguide dispersion curve falls between those two curves. Well, the pulse transmission, we knew was going to be associated with the group delay. So, we had to go in and find the second derivative of this waveguide dispersion propagation, I should say, second derivative of the waveguide propagation cost.
So, Felix would go through all that, but he hadn’t put in the material properties. I brought in the notion of bringing in the material properties. Well, we’ve long known that in the anomalous dispersion curve there’s a spot where the curvature reversed the sign, in the anomalous dispersion curves. On the one side of the point, it has one curvature. The curvature is the second derivative of the propagation constant, effectively. But if you look at, you know, solid-state theory and plotted the refractive index which is a measure of the propagation constant, if you will, in solid materials, it goes through this anomalous dispersion where the curvature is one direction on one side of a particular wavelength, or frequency, then on the other side the curvature reverses sign.
Well, where the curvature is zero, the second derivative is zero, and you would predict that the dispersion of the waveguide is zero at that point. And so, you would predict that there is an infinite bandwidth at that particular wavelength. So, this paper was associated with essentially finding out how the dispersion curve of the guided wave falls between the indices, the dispersion of indices of refraction of the two materials, the core and the cladding material. And it follows this nice smooth curve between them, and you look for the point where there’s zero point percent of curvature, and that should be a point of infinite bandwidth in the thing. And then I remember vividly that Maurer argued with us at length before we published the paper, but at that zero point, while it exists, the zero curvature exists for only one wavelength, one frequency, the delta function, and if you depart on either side of it why you begin to pick up a little negative curvature on one side and a positive curvature on the other. And, basically, all that says is that on one side of it, if I would launch a pulse of red and blue light into the fiber.
One side there’s zero dispersion point, the red pulse would travel faster than the blue pulse. On the other side of that frequency, the blue pulse would travel faster than the red pulse. At that point, they’re both travel at exactly the same speed. But, any source is likely to have a finite spectral breadth. We knew lasers didn’t have an infinitely narrow bandwidth. So, if you begin to plug in the spectra characteristics of the source, which we did in that paper, you begin to say, “Well, the dispersion will actually be finite,” it won’t be, I’m sorry the bandwidth would be finite, it won’t be infinite at that wavelength.” And we showed that it was directly proportional to the root square of the spectral width of the laser source. Well, the source that you were using in the transport. And that’s the first paper that anybody ever wrote on propagation characteristics in dielectric waveguides. So, I’m pretty proud of that piece of work that Kapron and I did.
Lassman:Okay. Let me ask one quick last question and then I figure if you’d like to break we could.
Keck:Get some lunch?
Lassman:Then we can start the next section.
Keck:Okay. All right.
Lassman:But, just a, this is more of a clarification too. In the U.S., in the United States at this time, with the Bell System, they control the market but you’ve got that, you mentioned that Chuck Lucy sets up these contacts with these other companies. So, Siecor, for example, the one with the joint venture with Siemens, so there is, in looking at the business outlook from the management side at Corning, the international prospects those are more, is that something that’s keeping, you know, Houghton and the other executives on board for fiber optics, given the limitations in the current U.S. market with the market control that AT&T has, or that Bell System has as a monopoly?
Well, there were a couple of things that I think kept Corning going. Certainly, those joint development agreements did, and they kept going through the late ‘70s. But the other thing was the patents that we had. We were building a very significant patent portfolio by now. Now, a little more history. When we moved into multi-mode fibers, another paper I’m proud of, when we moved into multi-mode fibers, if you had the, we were making a so-called step index single-mode fibers, where the core refractive index is constant across the entire diameter and then falls off abruptly at the edge. If you tried to do the same thing to multi-mode fiber you’d find that the modal interference from the many many modes that are traveling down the fiber would essentially bring the bandwidth from, these near-infinite levels down to something that’s on the order of 50,000 bits a kilometer, or something like that, bits per second per kilometer. You know, a ridiculously low band. But, papers had already been published on graded index lens materials — another interesting story. Whereby if you parabolically graded the refractive index of the material, in fact in the proper way, the parabolic fall off of the refractive index is the proper way, or was thought to be the proper way, then the, when the waves that travel, you know, all the way out to the cladding and essentially oscillate sinusoidally down the waveguide, when they’re out near the cladding they’re traveling at a low-refractive index media. So, they essentially travel faster.
The waves that tend to go more straight through the fiber, travel in the high-refractive index material and therefore travel slower. So, even though the path length is longer for the other rays, they travel faster in the region. And by tailoring your refractive index you can equilibrate those and essentially get the bandwidth back up to fairly high levels, and in fact in principal up to the material dispersion limits, if you’re very clever. Gloge, at Bell Labs, had authored a paper, a wonderful paper, where he had shown that the refractive index needed to be, have a square loss fall off minus delta. Delta was the index difference between the core and the cladding, fractional index difference, about one percent, .01. His theory said that the dependence of the composition as a functional radius had to fall off as R to the 2.01 power, essentially R squared. And, I looked at that, and studied that paper and said, “This isn’t right. He isn’t taking into account the fact that the dispersion of the materials are constantly changing.” And by this time, I’d hired another theorist. Kapron had left. And, I hired a fellow by the name of Bob Olshansky. I asked Bob to take a look at this, that I didn’t think it was right, he had missed a turn, or wasn’t differentiating properly if you put the wavelength dependence of the core and cladding materials in. And sure enough, Bob looked at it and added another turn to the series expansion and we found that Detlef had made a circa 20 percent error. And it wasn’t 2.01, but it should be 1.86 for germanium, and oh, by the way, it’s material dependent. Olshansky and I wrote a paper on the optimal alpha, the optimal profile for multi mode fibers and patented that.
Shortly after that, the interaction with Bell Labs ceased. Don’t know why. It just disappeared. I guess they — I don’t know. You’ll have to ask somebody down there. But at any rate, the relationship ceased and in nineteen —, and talk began floating around about dissolving the cross-license agreement. And in 1980, the cross license agreement was terminated. So, any patents that we got subsequent to 1980 were not covered by the cross-license agreement. So, this is before the break up. They clearly had access to a goodly number of patents but there are any number of improvements that we made subsequent to 1980 that no longer were a part of this. The other question you raised about the markets outside, that we intended to have, the answer is “yes,” but they were fraught with problems too. In this whole time period, telecommunications were still very nationalistic. The government ran telecommunications, and effectively didn’t want to let anybody else in. So, they had already told us that if we carried forward with this thing the manufacturing operation for fiber or cable or whatever was going to have to take place within the national boundaries of whatever country we were talking about. It was not yet clear that the traditional Corning model of “make a lot and ship it around” was going to hold true. But, nevertheless, we felt we had a strong patent position. We had these partners that were believing in us, and, you know, encouraging us to go forward.
I suspect that by ’79 the Judge Green sorts of issues were cropping up. So, that probably carried some thinking as well. Amo Houghton agreed to the capital expenditure to refurbish a factory we had in Wilmington, North Carolina, that was in the electronics business making resistors and capacitors, and bringing the fiber line in to there. So, we moved from the pilot plant down to Wilmington. We’re still making about 1,000 kilometers a year, was the volume. And, Amo is, of course, taking all sorts of flack from consumer products divisions and the catalytic converter divisions on, “Why are they spending the money on this?” And so on. And by this time, we’re sinking major amounts of funding. The hole we dug is multiple hundreds of millions of dollars, by the ’80, ’80 time frame. But, as you said, by ’82, or ’81 I guess, two things had happened. One the consent decree was through, but all the other telecommunication firms, MCI, Sprint, GTE, were seeing fiber as the way they had to go. And in ’81, I think, MCI was the first company that came in and gave us a contract for 100,000 kilometers.
Lassman:Am I correct here, that MCI had no technology of its own?
Lassman:At that point. I mean there was no infrastructure.
Right. They had to buy. They were busy buzzing around. And, of course, brought about the demise of the Bell System with their constant challenges that this was monopolistic. But here Corning is making 1,000 maybe 2,000 kilometers a year and suddenly we’ve got to ramp up the production to 100,000. Moreover, circa 1979, laser diodes had improved. Spot sizes were getting smaller. I had been working with British Telecom Labs, John Midwinter over there, interacting with him. Colleagues at different conferences, and had grown up in the technology together. Other fascinating aspects of this technologic evolution, British Telecom began espousing the notion that “lasers were good enough that we can get back to single-mode.” And so now the question that arose before us was, “what’s exactly the right design of single-mode.” We learned a lot more about the bending properties of fibers, in particular microbending, and things like that. And, so there was clearly going to be a trade off on core size to get light in, but if you made it too big the bending loss went up. And, we knew how those curves, in theory, went.
So, I posed a contract, a project contract with British Telecom, to design the single-mode fiber. And they didn’t take me up on it, but the Naval Research Lab did. My friend Tom Giallorenzi down there thought that would be important for the nation. So, we designed, came up with a design for a single-mode fiber. And, we were already to go and MCI came in and, I remember Dave Duke who was running the business by now, Chuck was an elder statesperson. The company wanted a young, energetic, entrepreneur, rising comer to take over the business. So, Chuck became his assistant, and Dave Duke took over the business. He had gone out to MCI and called back to the plant and said, you know, “I’ve got good news for you and bad news. The good news is they want to buy 100,000 kilometers at some price, and the bad news is its got to be single mode.” The plant was, all they made was multi-mode. In six months we transferred the design process from single mode fiber down to the Wilmington plant, and ramped the production from 1,000 kilometers up to a rate of 100,000 a year.
It’s the type of thing Corning had done over and over, and over, in its history. And did it again. It was just euphoric feeling at that point. That was ’82, when those orders came in was the first time that I’m sure Amo started breathing a bit of a sigh of relief that we knew we had a business. And then the Bell Labs break up, or not Bell, AT&T breakup now we got seven fiber-hungry baby-Bells all ordering fiber as well. So, all of a sudden the plants out of capacity and we’re just making fibers fast as we can. And the AR for the plant had called for us to make a half a million kilometers a year, at the peak of the business.
Keck:Took to, to the board to get the capital to spend on the factory. Half million kilometers a year. I don’t know what the volume is now. At the height of the telecom bubble people were deploying fiber at the rate of one and a half kilometers a second around the world, 365 days a year.
Keck:Yes. It is. You figure out what the volume is but it’s circa tens of millions of kilometers. So, there’s another indication of how big we thought the market was. I mean, we didn’t know. And indeed, it just mushroomed when people began to see the capabilities.
Lassman:Okay. The next section moves into your movement into management positions. Perhaps this would be a good time to break?
Lassman:For a few minutes?
Keck:That’ll work. That’ll work.
Lassman:Okay. [Recording paused.] Resuming on Wednesday afternoon. I’m going to shift the discussion now to your movement from research, in the research lab into management. This is in 1976, when you become manager of the Applied Physics department. Tell me the story about transition from scientist to manager, and the department that you took over.
Lassman:Inherited, in the mid ‘70s.
Well, I’m sure that by, even before then I had realized in Corning and then also in other laboratories, if you move from the bench to the management, it tended to have better opportunities for [a] higher salary. Ability to control destiny of your projects, and you can get people that can fill you in on projects and so on. And so, early on I had, we had yearly meetings, you know, performance discussion meetings with supervisors, and had probably indicated that I thought at some point I wanted to take the management track rather than continue on the scientific track. And, my reasoning at that time was, it was very simply that I thought of, you know, I had people working for me and I could chart a direction for them, why I could get more done, and more successes. So, all the way along I still, I guess I had in the back of my mind that I was a scientist, not really a manager, but that I thought I could manage as well as the next person. After you get into it why you find it’s not all that simple. But, nevertheless, that was what I had told people. And, Chuck Wakeman, Bob, later Les Gunderson, and Bob gave me an opportunity.
So, at one point as we were evolving the fiber, Bob divided his department into two subgroups, small, but you know little subgroups, and we had hired an electrical engineer, Frank Teal. And, one of the things we understood we were going to have to do in carrying this idea forward was not only advance the fiber work, but all the ancillary stuff that didn’t exist, we were aware that it was going to have to exist before a viable market for fiber existed. So, not only did you have to cable it, but you had to be able to connect it together, disconnect it, splice it perhaps, inject light in, transmitters and receivers, and then, oh by the way what’s the transmission format that you’re going to use? Oh by the way, what, you know, are we just sending telephone messages or, you know, is this going to be video transport, or what sorts of applications would one use? So, we had an electrical engineer that had joined Corning — now I’ve forgotten — somewhat earlier, and he was working on a copier project that we had, which is a total other story, but at any rate, Frank had an electrical engineering background and was a young up and comer, as was I.
And so Bob gave him an applications group and I had, I think, two or three scientists working for him, and a couple of technicians, and maybe five or six people. And, did the same thing with me. John Crowe was just hired. John came out of U.C. Berkeley and was interested in integrated optics. And, we thought, “Well, we’ve got an interesting new technology.” So, John came into my group. And, I don’t even remember now who else. But at any rate, I had two or three scientists that were assigned to me. Well, we were anointed research supervisors, or something like that. In other words, basically, they would try us out. Bob still managed everybody. But, nevertheless, on paper we were assigned management responsibilities. And that was the beginning of testing us out. And, apparently, we must have done okay, because I don’t know to this day exactly how it came about but I think Bob was really not happy as a manger. He was, he was a scientist at heart, as well. And, my history says that they anointed him the first research fellow at Corning, and created a new position in the laboratory on the technical ladder that was equivalent to essentially the, someone who reports to the director of research. You know, second down, but on the research ladder. I mean, this was a research fellow. IBM has research fellows. And sometimes they get accorded a vice presidential status, and so on. So, I think Bob was the first research fellow. And, it was when they made it in that position I took over as manager of his department.
Lassman:Of the applied physics department?
Keck:Applied physics department.
Lassman:So, that would have been 1976?
Keck:So, that was my first real anointing in a management position of any kind. Reported to Les Gunderson.
Lassman:And, who did Gunderson report to?
Keck:He reported to — Armistead, I think was still in charge. I’ve forgotten when Armistead became vice chairman. No. It must already have taken place. Jack Hutchins was probably, by that time, the — well today we’d call him — hmm. I’m trying to remember what the engineering division reported to Jack. They reported to Armistead. I don’t think it reported to Jack. So, Jack had the research and development functions. So, it would not have been the CTO title. I was going to say Jack was named CTO. But, I think they took it a bit more slowly. And, Armistead or someone else took the engineering division that had reported to Armistead.
Lassman:So, at what point did you start…?
Lassman:How does the breakdown begin between bench scientists, bench science and management, and how do you, how did you negotiate that?
Keck:In this time frame I was still, I still had a technician reporting to me, and said, “Sure. Go ahead. Do that.” And maybe it was, maybe it was only a year that I really had the technician reporting to me and then I had him report to someone else, and contented myself with continuing to look at theoretical aspects of propagation in fibers.
Lassman:So, you were still wholly engaged in the fiber optic realm?
Keck:Yes. My department was.
Well, in fact, by that time, yes, we had made a, that was when it happened. But, by ‘76, it was decided that we would create an entire organization devoted to optical fiber. Les Gunderson was to lead that. They dissolved the physical research directorate that he had had, which was sixty, seventy people strong, and in which Bob’s group sat. And Gunderson was to take those scientists and put many of them into the chemical research department. I think it was still called chemical research. But, he was to pick up some development activities as well as the research activities. So, Pete had a materials waveguide research group in the chemistry department, and we said we’re going to bring everything in to do the waveguides under one person. And, that was dramatically different than we’d ever done in the laboratory, at least as far as I knew. And then something that the department I guess, sort of a big project.
So, I managed the lab research physics part. Pete had the materials research. We continued to work closely together, and move whatever new fibers we were looking at at the time, largely multi-mode and trying to get the profile right. And, measurements of that, and so on. And, Gunderson also then had the materials development department, the coatings, fiber draw department, and we brought in, there were six of us. They brought in one, somebody from the engineering group to begin working at the machinery at which you’d need to move the thing forward. But, what the devil was the sixth group? I don’t remember. Well, at any rate. Gunderson was now running optical waveguide group. Lucy was building, well no Dave Duke by this time was involved. Chuck reported to him, and they’d begun to build up a sales, marketing, accounts receivable and all the functions that an operating division would have had doing this to sell fibers out of the pilot plant. But there was a third prong that came in, I guess a little bit later. It must have been.
The engineering group brought in more reinforcements. A gentleman by the name of Bob Ecklin. And, the Gunderson, Ecklin, Duke triumvirate essentially ran the business, all aspects of the business, total responsibility. Duke was, at the end of the day, everybody knew who was going to emerge as the leader of this business. He had grown the catalytic converter business, therefore he’d cut his teeth on, and had done a spectacular job of getting us some business there. And, now they brought him over to run this division. So, it, in my opinion, it worked out very well to put everything in one place and our communications, we didn’t have to worry about the physics-chemistry rift, we were both in the same group. It made just a great deal of management sense. I know they had outside counsel that came to bear on the equation, or — well at any rate, it was done. And, later on the activities of that group, and the success that that group had was used as the template for a process then we put in place across Corning, probably in the late ‘70s, early ‘80s — called the innovation process. We were the template. The reason and the fodder, the details behind that. Skip Deneka, who, he didn’t report to Gunderson. He reported to the plant. So, he may have reported up through Ecklin or Duke. Anyway, he had that innovation generation discussion point.
Lassman:What is it that makes that the model?
Lassman:Because, I’m thinking at that time period …
Keck:We’re still not successful?
Lassman:Yeah. I mean, that’s, you don’t have the — I mean there’s …
Keck:I may have the date wrong. It may well have been after the Bell break up and the factory started, turning. Don’t hold me to the date. It may well be in those Corning history books. Well, I still think it’s a most valuable process, and set of documentation associated with moving projects through research. And the whole notion, as I’ve said, so often is probably on that list. You progress only as fast as you can measure it. And we had a whole series of things that experienced people over the years said, you know, “You got to do this, and this, and this, and this, before you go to the next stage, before you can start spending a little more money on the project.” And so, there were gates and things that you had to accomplish before you could ask for the money for the next phase. That was a very valuable document, in my mind.
Lassman:Well, this sounds like, this is a much more formal process than what you described yesterday where — I forget who it was that you mentioned would go to Amery Houghton’s office and say, “I need ‘x’ number of dollars to do this. Fine, you know, on a handshake and you go do it.” Is there a transition from the executive, that type of executive management style, and also on the research level, someone like a Bill Armistead? He’s no longer, you know, in a, in a group.
This, this what I just described is that transition. Prior to that time, projects went ahead if Bill Armistead said they should go ahead. He really was knowledgeable both, accorded the knowledge both in the marketing as well as the technology side of the equation. So, Amory Houghton just relied on his judgment and gave him carte blanche. When I joined Corning, the adage, on the part of any sales and marketing person, they had to get to know, they advanced on the basis of how good their connections were with the laboratory, and how well they listened to the lab scientists, that counseled them on the business and how to proceed and so on. We were accorded a lot more power over, well and Armistead ultimately, in deciding, you know, big projects and so on. Then later came the past by the way.
So, it was almost a poor analogy, we were almost Mecca and the business people had to come up to Mecca and pay homage and then, you know, we’d tell them projects that we thought might be interesting. They’d go out and test the market and see if it was interesting, and begin to try and pull it out of the laboratory and so on. And, when did that, it was probably circa the time that Armistead got elevated to vice chairman. I was just a senior risen leader. And, Jack Hutchins, new kid on the block, and the general managers in the line divisions, that are making the money after all. A lab is a drain. One in particular, Bill Hudson, reportedly, thought that the balance of power needed to change, and began the notion that the research group shouldn’t just have corporate funds but they should get half of their funds from the corporate coffers and half of it from one of the operating divisions, so that there was a better tie between scientists and the operating divisions. I think the scientist had to do a project for an operating division or they didn’t get their project funded; well they’d be a lot more interested in listening to the divisions and what made sense from the marketing standpoint, and so on.
So, that was, it was when Armistead got elevated that the shift began to occur, where the operating divisions began to have much more say in how things progressed. And, that evolved to, then, this well orchestrated innovation process. But, a group from both the divisions as well as the laboratory and engineering got together, and said, “Well,” — and the beauty of that innovation process is that we didn’t just set up measures that the laboratory, with the technology group, had to complete before it could move to the next phase of development. The marketing people had a series of milestones they had to meet. The manufacturing folks had a series of things that they had to meet at each of five stages in the evolution of a project. It all came because this triumvirate of Duke, Ecklin, and Gunderson — everybody successfully ran the waveguide project. They just watched what they did and wrote them down, and that became the basic innovation system at Corning.
Keck:I’ve later found many companies, Kodak, has something very very similar. They had four stages in their evolution. We had five. But, you know, they’re all stage-gate, sort of, process.
Lassman:Before this transition you said that corporate was the primary source of funds for the lab? So, would that get, Armistead would get that and he would just parcel it out?
Keck:We never knew exactly how that happened. It was a lot like I described earlier. Tom Hart, you know, talked to Chairman Houghton, and Armistead was of that yoke where…
Lassman:Then, in this first management position you had in 1976 to ’86, you’re doing science, heavily involved in the science? When…?
Well, it’s in that, in that time period. From ’76 to, I’d say about ’81 or ’82 I was still trying to be reasonably up on the science. You’re reading a lot, doing some theoretical calculations, and working in detail with the scientists and so on. I think it was ’81, Dave Duke was asked to give a paper down at an SPIE conference in Washington, I believe, and I was as well. So, here’s the business leader and the lead scientist. We went out to dinner. Dave asked me how, what it looked like by this time we were going to get large orders out of the telecom companies, that the break up would occur, and so on. I’m pretty sure it was ’81. He basically said, you know, “What do you want to do next?” And, I began to describe how I thought that there was an additional market that Corning should participate in associated with, the transmitters and receivers, connectors, where we connect the fibers together. Somewhere I had gotten the notion that we were going to need splitters in the network, things attached to the fiber for signaling to go on your multiple beams. And basically began describing that there’s a set of componentry that I thought Corning should begin to investigate, and that I wanted to be in the research group in that direction. I’ve forgotten whether Dave said, “Okay. You got a million dollars to do that.” I think he did. Or, Duke told Jack Hutchings, the leader of the R & D, that he should, had an extra million dollars that Dave was going to siphon off the profits of the business and give back to the laboratory for me to hire some additional people into the fold. And, that money, that never trickled down.
So, why I don’t know, but it didn’t. And, when did they do the optics strategy? I guess it was ’84, I think. Yes, it must have been ’84. Tom McAvoy was still president. Somehow Tom McAvoy got this notion that I should propose an optics strategy for the corporation, and the word had floated back down through Jack and the fellow I was reporting to at this time. They had dissolved the laser guide group by early ‘80s. Well, at any rate, I didn’t report to Gunderson. I reported, now, back to research division. And, Harmon Garfinkel was the gentleman’s name. And, McAvoy asked if they’d get me a group together and enunciate an optics strategy, something to follow on beyond optical fiber. And so a group of us got together, and it was a bit of a disparate group. I mean we had a whole hierarchy, there was a steering committee and it was going to steer us, and we were the first results and then they were going to take that ultimately to the management committee of the corporation. And, the six people that ran the corporation. And so I worked for — and Schulz had left by this time, I think. They sent Pete away. He was, he was on a fast track, management track, but just thought the grass was greener elsewhere. He’d gone to one of the advanced MBA, accelerated MBA programs at Harvard or MIT, and the company had sent him to that. He came back from that and I remember he came into the office and closed the door, and said he was leaving. But, he went to one of the small startups that were beginning to populate the arena. Well, at any rate, at that point the corporation started getting worried that I’d leave, and started putting the golden handcuffs in place.
Lassman:Were you thinking of leaving?
Keck:No. But, at any rate, they began to accord me a bit more visibility in the corporation, and so on. So, it must have been ’84. They asked me to pull this optics strategy together. And so I gathered a group from a couple of the divisions, a couple, out of the marketing folk. They basically told me who was on the committee. It was a new business development director from one of the divisions that wasn’t doing very well. That was one of the people. I remember. We got one of his senior marketers to do a lot of legwork for me. And, so, it was a bit of a ragtag group. Somewhere along the line, I got the notion that we’re in the information age, and what did you do with information? You generated it, you stirred it, you processed it, you displayed it, and you transported it. I don’t remember whether the optics cross is in here or not. But I came up with this cross. No, it’s not. Unless it’s in the back end of things. I don’t know, there’s the one I was — the waves. I don’t see it. Nope. Not there.
Keck:Well, at any rate. It basically was just a cross and you know circles on each end identifying these things. And then we began to talk about how Corning, in total, we had some optical disc work going on in the laboratory, trying to make glass disks for memory devices. In the processing arena we didn’t really have anything, but I suggested some opportunities in nonlinear optical materials. And we had a material that was thought to be highly nonlinear. A semiconductor glass. This display obviously had the cathode ray tube, because it was still — flat panel wasn’t on the screen. Although there were programs in the laboratory beginning to work on flat panels. So, we lumped that in. What we have in the generation? We didn’t have anything. But then I put in the passive componentry, and the hanging lens of fiber would connect the fiber to these — excuse me a second. So, I added these little bubbles at the ends of the fibers and began to describe how optical amplifiers, there were papers beginning to flow on those. I had the connector. Siecor was already doing work in the connector area.
Lassman:Siecor? Siemens Corning.
Keck:Siemens Corning joint venture. I thought we should be doing something laser diodes and transmitters and receivers. And, I put that forward. And, connecting, couplers, I thought, I had some ideas on how to make couplers. And so we, I began to sketch out some componentry that we could make out of glass, that looked a lot like fiber and would hang on the ends of fibers and enhance the distance, and so on.
Lassman:Are your customers also asking for these components, too? Is there also a market pull for these things, or are you conceptualizing these new technologies beyond the fiber optic cable as just part of a wholly integrated system?
Keck:No. A lot of it is coming from this joint development agreements that had occurred five, six years earlier. The customers had, were working on themselves. And, in some areas it was, well I thought Corning had a better shot at making some of the devices, like a fiber coupler. I thought we could do a better job at mass-producing couplers than the things that were out there. Some of them were out there, that were out there but it was small companies that hadn’t yet achieved a toehold in the marketplace. And, a lot of it was just one off in small companies. It had been a little bit of this, and it was all hand done. No manufacturing capability. Well, at any rate, I enunciated this, my steering committee was largely lab folks brought on to it. And they had brought me into the management committee. And, this was Jamie Houghton, by this time. So, he took over in ’83.
So, yes, ’84. I didn’t have the date’s right. Fred Ackerman is running one of the divisions. I’ve forgotten which one. Marty Gibson was running the life sciences division, or was in the process of doing lots of acquisitions in this time frame. Tom McAvoy was the — Tom had relinquished his presidential position and was this chief technology officer, essentially. So, he had both the laboratory and the engineering division. And, I’m trying to remember — I guess Jack was still in there. Jack Hutchins was still there at this juncture. They brought Dave Duke in from the division, but there was someone else that — I don’t think Dave was one of the six-pack. Oh yes. And Dick Dulude. Dave Duke reported to him in the fiber business and there was one other business that reported to Dulude.
So, those were the business leaders. Oh, and then Ben Campbell is the CFO, McAvoy, and Jamie. That’s the six-pack. So, those were the six. And, I had to pull together this little forty-eight slide package of this optics cross and the order I asked for — I showed them all the technologies we had that fit. There were so many devices that we could make with these technologies. And if I added a few more technologies, we would have made even more, and more. No it was the other way around. Device relevance. Oh my word. I’ve forgotten the chart. Oh, forget it. At any rate, there were a couple of masterful charts that I felt very good about and obviously, they did too. And I asked for the order at the end of it all, I said, you know, “Corning has every right to play. It’s and $11.8 billion market.” Some of the marketing folks had pulled all this together, and I included all these things. And, we had to write the play. I pointed out that we had some missing technologies and we needed to bring them. And so the strategy was, maintain and advance the ones we’re either neutral or ahead of your competition’s technologies, or acquire the ones that you didn’t have. One of the ones I said we didn’t have are organic chemists, and organic materials capability, and we needed that for the polymer coatings on fibers and things like that. Get rid of any technologies where we’re driving ahead of the competition. Let’s not waste any more money on them. And, I said, “And, oh by the way, I need eight new scientists to facilitate this strategy.” I’m still a bit naïve at this point. They said, “You got the order, and go for it.” It took two years before I got the money to hire my other scientists. But, gradually I began to hire a few people. I hired Doug Hall. And, he came in from Lawrence Livermore. Had done a lot of laser glasswork out there with a well-known name in the glass laser business. And, he began to do a project in optical amplifiers? Optical amplifiers. Which is a business that, as we grew the photonics was the bread and butter of the photonics.
Lassman:The optical amplifier, what’s that? That’s not  in the traditional communications system?
Keck:Well, a repeater. Well, we had repeaters in the optical communication too, but what you had to do was, the optical signal came along — oh, what were we running, 600 megabits, whatever, in the early days. You’d have to detect the signal, clean it up through electronics and then remodulate on a new laser diode and send it the next length, typically forty or fifty kilometers was all you could go before you needed to repeat it again. And, if the bit rate changed, or if the wavelength changed, you basically had to rip out that repeater and put in a different one that would be sensitive to whatever the new wavelength was, or have the right laser diode for the new wavelength on the other side.
Lassman:So, and the wavelength would change depending on what type of instrument?
Keck:The source. Whoever supplied your source. I’m anticipating things a little bit because WDM — well I have a patent on the WDM device back in 1971, or two, but WDM technology, I mean we come into the forefront until in the mid ‘90s.
Keck:Wavelength Division Multiplexing.
Keck:When you put a whole bunch of different colored lasers all transmitting through the same fiber and then split everything out at the other end.
Lassman:So then, you don’t need to trans…?
Keck:Right. If you’d had the electronic, electro optic repeater each one of those wavelengths would have to have a different detector, that filter, all the rest of them. would avoid it. And it was going to be a nightmare. If you can amplify it, you just put the thing in. It doesn’t care. It just amplifies whatever wavelength comes through. It was beautiful. So, at any rate, Doug started a program with me. I guess I hired him probably eighty — well, when was I crowing about — I think I said Doug made the first amplifier in the laboratory in ’87. Ninety-seven is ten years. It must have been earlier than that. Well, at any rate, the amplifier business almost exactly paralleled the evolution of the fiber business, as to when we built the plant, when we got the first big order, when we made money. So, I’ve got the amplifier date so wrong, I may have them on a slide in there. But, at any rate, it was the enunciation of this strategy to move forward smartly in the telecommunications arena, and in the optics arena, in a broad sense. But, I put forward maybe four, and they named me a director in ’86. Then, well, I continued essentially in that business arena until I retired entirely.
Lassman:Well, when you move into the director of optics and photonics research at what point are you starting to do less and less science?
Keck:I’m out of the science, scientific arena by probably ’81 or ’82.
The last paper I published was a Journal of Laser Technology paper that I did with Dan Nolan. And I remember that so vividly in that it was one of the trips that I made back to Michigan. Too far. No, actually, I can’t say that. Because we really didn’t publish it right away. I’m sorry. I don’t see what I’m looking for. Hmm. I haven’t updated [my publications list]. Well, actually I can’t say quite so definitively. Well, I don’t see what I’m looking for. There’s a trip to Michigan there that was, maybe, I don’t know, Christmastime, or whatever. The family and driving at two o’clock in the morning. And, in my mind, I was going through a set of equations that detailed the mode coupling properties of multi-mode fibers. This must have been — no it was earlier.
So, it was probably still in the — we started the single mode design in ’79, and so there was a period of time where we were still doing a lot of multi-mode work that was going down, while single-mode increased in effort in the group. And it was probably as it was winding down in the early ‘80s. Longer and longer segments were of interest. It turns out if you did mode coupling I said that there was a dispersion problem with the many modes and a multi-mode fiber. No matter how good you made the gradient, there was still all this residual dispersion we never could get around. So, one of the ways that was conceived of early in the game, was that if you transferred power from one mode to the other, as you traversed it, there was a random lock problem, if you’re familiar with that? If you were, this position, yourself at a given point, and randomly walk in steps, randomly in any direction, so you head in a direction and from that point randomly go in any direction and from that point randomly go in any direction, statistically you would be within a circle the square root of the number of steps away from the step times the step width, or step length, from the starting point. So-called random walk problem.
Well, at any rate the mode-coupling problem was the same sort of thing. And, I worked out what the final pulse-broadening equation ought to look like and then came back, and Dan Nolan was my theorist at the time, and he’s still at the lab as a research fellow, and conferred with him and we got it down on paper and published a paper on that. And that must have been — I thought it was in the list there, ’83-ish. But it may have been that we did the work and looked at it internally and didn’t publish it until quite a bit later. I’m pretty sure we published it. But, but my modus operandi changed. I didn’t get involved in the details, but I guess I still considered myself an inventor, or somewhat creative, and so I’d suggest to scientists “what if we tried something.” And one of the “what ifs” I tried with a young scientist that we hired, Venkata Bhagavatula, was I asked him, “Well, you know, we grade a multi-mode fiber up, what would happen if we graded a single-mode fiber?” He put some sort of shape to the index, the composition profile of the fiber. He invented a class of fibers that I named segmented core, Segcor fibers for short. They are in the list here someplace — wherever the Bhagavatula papers are. I saw it just a minute ago. Olshansky has left Corning by now.
Keck:Yes. Segmented Core Single-mode Fibers at Low-Loss. But I published it. He did most of the work. I suggested the idea, and Walt measured some of the properties of…
Keck:This was, it turned out that the zero dispersion point that we had discovered way back in ’71, in silica, in the germanium-doped silica it ought to have occurred about 1310 nanometers.
Lassman:This is your  and this is the modes?
Keck:The point where the, all wavelengths traveled at exactly the same speed in the fiber, independent of — it depends on the source spectral loop, but the bandwidth of the fiber, essentially, goes to infinity apart from the finite spectral loop of the source. So, that was the point at which you liked to operate. But, in the intervening reserving years we had gotten the water levels in the fibers down now 500 parts per, 5 parts per billion. And so the attenuation that those overtone bands, harmonic bands, that I talked about earlier was essentially nonexistent. So, there’s no band at 1,300. And so, you could continue out the Rayleigh scattering curve and it’s still going down the wavelength to the fourth power. So, it was lower-loss region out of 1,550 nanometers is the end, end result. And then there’s a huge absorption that starts simply from the silica matrix vibrations that occur, and it’s intrinsic to the silica glass. So, there’s this beautiful low-loss window at 1,550 nanometers. And so the question is, “Could you come up with a new waveguide design, a profile design that would shift the zero dispersion point from 1,300 out to 1,550?” Because now you’d have the lowest loss and the infinite disperse — or infinite bandwidth occurring at the same point, which is clearly what you want. By then dispersion shifted, fibers talked about by some of the Japanese scientists. They had used a step index — Well, no I guess they had a “W” profile, Kawakami and Nishida. So, Bhagavatula had played around with the light gradient of the core refractive index could be — we knew that the profile effect of the microbending losses that the cablers would suffer if you did it wrong.
Keck:It was again an optimization trick of shifting to the zero dispersion, minimizing the bend loss, maximizing the size of the core so that you get light in, and so on. And that whole class of segmented core fibers resulted from that. And, Corning over the last — well, during the bubble, had invented a variation on that theme that we called the LEAF, Large Effective Area Fiber. We found a way of making the core even bigger, and still maintaining all the rest of those properties. And that was, and I gave a paper at some business conference. It was the highest sales product in the company history. It had scandalous amounts of money. So, all along we, I contented myself from the technological standpoint of suggesting things to scientists and some of them they picked up, some of them they didn’t.
Keck:I had some up with a coupler scheme. Eventually I called multi-core, and one of the marketing guys wanted “multi-clad,” and he was running the show. So, we had a multi-clad coupler product out, and Dan Nolan helped me pick that one up and a then a scientist built a machine that would automatically make the things, in my lab, and they funded that. So, that was more the operation that I personally did to get my kicks out of it and still continue with some science using, you know, scientific intuition as opposed to, you know, the great experimental procedures If the experiments were run why it was one of the other scientists that ran those and, you know, reported the data to me or something like that. But, I didn’t actually get involved in the lab, and do the experiments, or anything like that.
Keck:But, that’s why my name appears on some of the papers later on.
Lassman:Sure. I, so now in the ‘80s, as you moving, you’re — well take in this case director of optics and photonics research. You’re in this, you’re in this management position, and I want to shift the discussion a little bit. Now that you’re taking a more global view of the fiber optics field, at this time the company is, has gone through some corporate diversification, and is in the process of diversifying further, and I want to get your views on that, and also about larger changes in the business environment in the 1970s and ‘80s. You talked a little bit earlier about the change in research strategy, from Armistead’s time to this new innovation process, and so I want to broaden that out a little bit and talk about as you’re taking a more global picture of the fiber optics business and the markets, how do you connect that to the larger business environment and the company’s efforts to diversify?
Well, yes, the company was clearly diversifying. Jack Hutchins, one of his hot buttons was the whole lab technology area. He launched a substantial effort in that. I had some people, although there were scientists who were working at trying to see where the physics would fit in this new arena, this biotechnology arena. And I had the notion that physics could play a role in measuring things and understanding what the heck was going on. Remember, in the optics cross I had also talked about generation of information. I said, “Our key to generating information was in building sensors of one form or another.” And so, we have a, one of the things we hit on in conjunction with some of the biotechnology group that we hired was a fiber optic fluoroimmunosassay using the evanescent wave in the fiber to excite fluorescent back scattering from organic molecules that were in proximity to the fiber. And then monitoring the fluorescent return. And the question is, how do you get the fluorescent return? Well, the notion was you’d immobilize an antibody on the surface of the glass. And Corning had patents on how you immobilize enzymes and biological material on glass.
So, we had some patent strength there. And the notion was that these antibodies would be receptors for the antigen that you were looking for, in some milieu of organic matter. And so you dipped the probe into the stuff and you slosh it around and if some of those unknown species were there, they’d bond to the antibodies. These things had been tagged with fluorescent tags. If we can excite the fluorescence and they got close enough to the fiber the evanescent wave didn’t get too far from the fiber, so we weren’t looking at fluorescents in all the goop. It was only those molecules that were bonded to the thing. We almost had a fluorescent sensor that had extraordinary properties relative to radio amino assay, which was the technique at the time. It was out and about. As the company is beginning to do these other things, all of us were being asked to look at where it might fit in with something else. And I was ready, willing, and able to do that with the physics group and still keep it in the context with a photonics evolution. It still looked to me like I was remaining true to the strategy that I had set for the corporation.
One of the quotes on my sheet, “the collision of creative minds from different disciplines that produces the biggest opportunities.” Pete and I worked together, chemist-physicist. And so, I was firmly of the mind philosophically that we want to get the biotech and the physics people together and see what we could come up with. But, contrary perhaps to — and I guess I had a hand in pushing in that direction — the, a lot of the early industrial research, that was in their 30s, 40s, 50s, and so on, was of the notion you do basic research and all good and wonderful things will come out if you do it long enough. And, I was more of in a mind that a prepared mind, if exposed to the problems that are out there, and if you select your problem area, on the basis of the way you think, there’s a reasonable market opportunity; you bring the creative minds in to do your work in a particular thrust area. And, oh by the way, the trick is to define the problem.
Odds are that creative mind is more likely to come up with a solution faster than if you simply said, “Well, we want to go in the biotechnology direction. Figure out something to do.” So, it was moving more to an applied research. And, by the way, Bob changed the name of the department from fundamental physics to applied physics in ’73, ’72 or ’73. So, there had already been a notion that we had to get more applied. And, my thinking evolved out of some of those things to the notion that what we really needed to get back to the point where we were interacting more with universities. And, letting the universities do the basic research, and then for us to be clever enough, know the problems, know the problem areas, and connect to them. We’re the agents that would put two and two together and pull the whatever the basic understanding from the university research was, and modify it and expand upon it, but apply it to a particular problem, and begin to move in that direction.
Lassman:This is when you’re in this position of director of optics and photonics?
Keck:Yes. By the time I’m director of optics and photonics I’m sure I was enunciating the notion that scientists had to get out — I had worked very much with the Institute of Optics [at the University of Rochester], and Duncan Moore, my friend up there, and you know, see what programs he had and we had joint work, and I helped fund some of their stuff. And, it was a source of students, plus it was a wonderful source of, well, new ideas of how to measure things, how optics worked, and what the forefront of optics was doing, and things like that. One of my laments was I never could get the scientists out of the lab as much as I thought they needed to.
Lassman:I want to know what was in place before then, in the research laboratories, about the nature of what kind of research should be, what the researcher should be doing? Was there any…?
Keck:Where the ideas originated?
Lassman:Yes, was there any sense of university collaboration?
Keck:Yes. I won’t say that they didn’t have interactions with universities.
Lassman:And how was your strategy received among the executive leadership of the company?
Well, the latter I can answer real quickly. It took a long time, and you had to say it over, and over, and over again. You know, I catch on real quickly, just tell me six or seven times. This was sort of the way it is. And, the same with the scientists. When the scientists, well for the most part, well remember we grew up when Gale Smith was the liaison was going out and canvassing the world for new ideas. So, apart from leading the scientific literature, which frankly the scientists, some did a lot and, but most of them you didn’t do a whole lot of. That was another surprise to me, I think, when I got to Corning. Most of them eventually became enamored with their own knowledge base and they didn’t pay a lot of attention to the literature. Who else was doing stuff? Well, that’s a little bit of an overstatement. And there were clearly some that spent a lot of time and knew the literature forward and backward.
But, my sense was that as the photonics wave and telecommunication, fiber telecommunication wave, started there was a whole new learning on the part of, well the whole industry. You know, the journals changed where we’d publish things. The people with whom you interacted changed. Conferences sprang up and you began to be a whole new cast of scientists, the sorts of scientific endeavors where it changed from old Corning. Well I’m trying to remember whether, what university interactions we did have. As I say, it was largely one person. A few of the managers, well Charlie Parker and part of the analytic services group, physical research, and he introduced me to the University of Rochester. So, he clearly, had interactions with the university. But the scientists by… But, it just didn’t seem to me that there was a lot of traveling taking place, and as, well, my only benchmark, then, was as the telecom arena started and my conferences began to spring up in that, I began to go to teaching every one, because there would be something brand new that you had to bring back.
It was just a fast moving field. Probably, part of it was that glass science was probably pretty slow moving field from the technology standpoint. You know, I later charted the so-called technological momentum, and tried to make sure whatever effort, whatever effort we had in Corning, matched what we thought the momentum, collectively, in groups of us in consensus, would emerge as to which pieces of technology we thought were the ones that other companies were really pushing. And, there’s a lot of technological change. We wanted to make sure we began to address those. So, it was near thinking that came in as to how you wanted to conduct the research and how you followed things, and where you put your efforts. And, all that sprang up. We wrote the book as we built the technology, in many respects. At least for Corning. Perhaps other laboratories had long since learned these. But, for us, I think it was, we were probably learning it for the first time and shifting the corporate culture quite a bit. And, well, as the field evolved and, you know, we saw the sizes of other company efforts in this arena, we couldn’t match it. In some sense, that’s where the notion, in my mind, began to gel that we had to look outside and work with the universities to come up with new ideas. Now, the bad news on that strategy was that universities, in order to garner funding, were becoming more and more applied themselves. And so, their opinion on this was shifting over. You found more of them beginning to start their own businesses, coming out of universities. Universities were encouraging, beginning to encourage intellectual property and things like that. All of a sudden the group that I hoped was to be the genesis of possible new ideas and new pieces of technology were beginning to be in competition, in some sense, with you. They clearly didn’t have the resources that we could amass and put on a project, but it was, I recall thinking many times, you know, “Was this the right strategy that was moving us to a much more applied research focus?” And, departing from the basic research sorts of things that you did.
Lassman:Was there a resistance to that at all?
Keck:Oh yes. The scientists in Corning were steeped in the basic research tradition so clearly the materials folk were adamantly against that move. Well the pendulum has swung back, I’m sure. The fellow that is now director of research at Corning, David Morris, I guess I shouldn’t be making comments, because I really don’t know. He used to work for me, and he was very much on the side of doing basic research. So, my guess is that they’ve shifted back to doing a lot more of the basic research. And we’ll see whether they generate the same number of marketable ideas. At one point, I plotted my group’s batting average. We were batting about 400. Of the projects that the group worked on, forty percent of them made it into production. And, typically, you’ll hear corporations saying that out of 100 ideas only one will get into production. Well, we clearly had an advantage and here’s a burgeoning new field, and opportunities abounding everywhere, but we still picked and chose some ones where we could bring our technology to bear, and where we had the competitive advantage and knew more about the thing than somebody else did from whatever source. You know, the Er-doped fiber optical amplifier played to our strength and made from a special erbium-doped fiber. It was a fiber device that we should own. But when he had to buy the laser diodes, and I was now out of our strength. And that’s when they began, the business groups began the acquisition of outside technology to bring to you. I would have been happier if they’d given me the grow it internally, but the equations at that time when the bubble was going up said, “You can’t afford to grow it. You got to buy it.”
Lassman:Well, I want to stay on this topic. When you go to director of optics and photonics research, that’s in ’86, and you’re in that for a decade. And during that period that’s when the bubble starts to grow. Tell me where things stand in 1986 in terms of the fiber business and the components business, and then what the progression is, and this change from internal development to acquisition.
In ’86, I’ve presented my optics strategy in ’84, and that was in large part what allowed Dave Duke, who had been running the fiber business, he came back to the laboratory as the CTO, and he named me director. So, all of that stuff led up to my ascension to a fairly responsible position in the corporation. But, about 1986 the fiber business started to go in the tank. If you looked at the deployment, the long haul market had, to a high degree, been taken care of. And, I remember, you know, we were still practicing what we had learned in the innovation process. So, I told you, we had the business manufacturing technology groups get together, we still did that. And, so, I was part of the technology triumvirate now that met with the leader of the business, John Suwinski. In those days, I forgot who was in manufacturing now. Oh, Bob Forest. And, so, we’d get together and chart the strategy of the business. And, I can remember a strategy session in New York City where we were trying to figure out how the devil — we thought fiber to the home, all the hoopla coming out of the R-box was that the fiber to the home would start circa 1990. And here we are ’86.
The fiber business is starting to level off and decline a along the long distance single-mode of fibers. And, how were we going to bridge from ’86 to 1990? And, get through that four years and we were projecting the business would go down, and we might have to reduce staff, and certainly reduce the project size, and shift people other places. The business did decline but the corporation sustained us through that. We were able to make enough arguments that we shouldn’t shrink too much. We didn’t hire anymore. We bridged that. We made our first amplifier, and in ’87 I enunciated a follow-on to the optics strategy, the one I’d done in ’84, and we fleshed that out a little bit more. I didn’t have near the audience that I had with the first one, and we got into a lot more detail than the first one, and it lay there. But we enunciated that we should have — well, by this time we’ve created PCO too, the laser company that I mentioned that was ultimately a joint venture with IBM.
Lassman:Could you just maybe briefly talk, for the purpose of the tape…?
Yes. As part of the optics strategy I said, Corning needed the semiconductor laser technology, the transmitters and receivers, and eventually pumps for our beam amplifiers. And, a friend of mine had started a fiber optics systems communication course in 1971. The University of California, Santa Barbara, in the summer. Mick Barnoski was his name, and he came up with idea that he’d have these two-week summer courses up at Santa Barbara, and he’d bring in scientists from all over to lecture at this thing. We’d do a two-week advanced technology sort of seminar. And, we started the first one in ’71, and ran that for sixteen years, every summer.
So, I’d go out and do two days on fiber and fiber propagation, and Mike would do some stuff on connectors. And, Henry Purcell from RCA had a talk on laser diode labs. By this time, he was in the venture capital community. Who else did we have? Well, we had a variety of others that participated in the course. That, by the way, was an important aspect in the whole evolution of the technology because in 1971 one of the things we realized that there were very few people that knew anything about (A) a fiber, and (B) about optical communications. And, if this thing was going to go anywhere, we were going to have to teach this. I mean, colleges weren’t teaching it. They didn’t know anything about it. So, we started this two-week course, and then others sprang up around the country. But, ours was the premier and as I said, we ran it for sixteen years, and I had a wonderful summer vacation every year. I went out to Santa Barbara and laid around on the beach and then do two half days of lectures. And, they paid for it.
Lassman:That’s an interesting point, about how you disseminate this new knowledge? Are there textbooks that are coming out in this?
Keck:Well, we wrote one.
Keck:We took our lecture notes and basically put words around them and published a textbook that Academic Press told us was a bestselling book in the ‘70’s. Well when did it publish? I don’t remember. And, there were two printings of it. We did the first one, and I think they were up in the tens of thousands, which, as I was told for a textbook is pretty good. And then we put out a second edition, updated it in ’80, early ‘80s time frame.
Lassman:So, it was in print for a long time?
Keck:Well, it probably still is. I say that, I guess I don’t know. I guess I don’t know. I’m sure it’s been supplanted by others.
Lassman:And then corresponding to that did there begin to appear special programs in optical technology in universities for specialized training? Or did that remain in physics departments or the existing science departments?
Keck:Mostly engineering departments. Some physics departments. My guess is that a lot of them would offer specialized stuff in physics and materials, or something like that, and they taught about laser diodes, and emitters, and things like that that bordered on it, but the mainstream courses on fiber telecommunication eventually evolved into the electrical engineering departments of schools. And, that was an interesting trend. The physics faculty were slow to jump on this wave, than were the engineering schools. On the whole optics, and optoelectronics arena, you began finding a lot more of the basic work was going on in electrical engineering departments at Stanford, and MIT, and places like that than were going on in the physics departments.
Keck:What’s my take?
Lassman:Yes. Take on that?
Keck:Well, I don’t know. It made, again, come back to some level of arrogance on the part of we physicists. That, you know, this is engineering stuff. It’s commodity stuff. There’s nothing new to learn here. It’s not basic enough. Well, for whatever reason, you know, a lot of the new stuff and the amplifier technology, and switching technology came out of engineering departments, electrical engineering departments, more so than came out of the physics departments.
Lassman:One other, I’ll ask one other question. While we’re on this topic. Is the fiber optics program is ramping up in the ‘70s, going from just your three-man team in the early period? I want to ask a question about recruitment. Given the lack of, that there aren’t schools turning out lack of textbooks, a lot of the trained constituency, that can work in this field, how are you getting people to come work for you, and is there intense competition for experts in this field among your competitors?
Keck:Well, the reason I’m pausing is during the dot com, the bubble, there was intense competition.
Lassman:I would imagine.
Keck:That, frankly, drove us up a wall. But, I don’t recall that there was that big a problem in getting people. But, we recognized that we weren’t going to be able to find somebody literally trained in the particular area. We just had to look for somebody that was bright, and could learn, and then bring them in and begin to hand them the textbooks, or the papers, or whatever, and get them up to speed. Bob Olshansky was a case in point. We needed some theoretical help, and he had just lost a job at CERN in Europe. I guess he was a postdoc or something like that, or studied at Brown University, doing SU-3 symmetries. I mean, esoteric stuff. And, I hired him, and he was just a quick study and came up like that. Dieter Marcuse, at Bell Labs, had written a book on mode coupling and waveguides, and Bob read through that, and the next thing I knew he was writing out theory for graded index fibers in mode coupling. It was very, very complex stuff. It had never been done before. So, he, he was just a remarkably quick study in the area. In other areas, I mentioned we hired Doug Hall, as I had identified that we needed more people working in the area. Doug we picked up as a laser jock out of the Lawrence Livermore Labs and he had been doing glass lasers, so, you know, whether it’s in a hunk of glass or it’s on a fiber it didn’t matter to him; he knew the laser physics. So, you could find people that were close enough aligned to things that you needed, and then teach them, learn telecommunications with all the rest of us. Because we were basically writing the book as we went.
Lassman:As you went along?
Lassman:All right. Given that, let’s get back then to where we started.
Keck:Where are we?
Lassman:Leading up to the joint venture with IBM.
Okay. So, this friend that I taught with for sixteen years had let me know for several years that he wanted to start his own company. So, I came back to Corning. Les Gunderson had left the lab and the waveguide stuff and decided that there were more career opportunities in the business, that he had to punch his ticket in the business arena. And so, he’d gone down there as a new business development director in the fiber group. So, it was going to be his job to work with me to bring to market my fiber optic, my photonics optic strategy, dream, if you will. And so, one of the first things I suggested to him was, “you know, we needed this laser diode, and why don’t you go talk to Mike Bernoski and let’s see if we can have him build us a laser company.” So, he did. And we hired Mike on retainer to develop a business plan and got him introduced to all the high muckety mucks at Corning, and got their blessing that, “Yeah. We were willing to back this young fellow.” And so Mike started his company, and he was from California, and set it up out in the L.A. area. We had hooked him up with Plessey to get the laser diode work. We had still joint development interactions with them. This was circa ’85. So, Mike got his company underway. About this time Plessey rolled over, started having economic problems, got bought by a German company. Don’t remember.
At any rate, they were going through their doldrums and I guess we got a license to some of the technology, but technology transfers was no longer possible. Or, maybe we already had enough. So, all we needed was the license. I guess that was probably more the case. At any rate, we still needed a partner. Corning didn’t want to go it alone in this thing. So, Dick Dulude had relationships with a high-level person at IBM, began the discussion there, and ultimately IBM bought out Plessey’s share of the joint venture, the company that we created. And they asked me to join the board of the company. So, Jim McGroddy became the CTO of IBM, and he had one of his accountants, whose name I’ll come up with, and then Corning had, I guess, four people on the board. Two from IBM. I think there were five of us. Well, at any rate, it was a wonderful learning experience to watch a small company grow, and yet try and marshal a degree to which the big company mindset and overlaying role of the bureaucracy that a big company needs to have, putting that at the proper level on a small company was an interesting educational experience for me. And, we ran the business for five years and IBM began to have a downturn, circa ’90, and I’ve already said that the fiber business was soft all the way through, through ’86 to ’90. And, so we were basically, at that point, closed the doors on PCO.
Lassman:That was Plessey Corning Optics…?
Keck:Optoelectronics or something like that. It was eventually, it started out as PlessCor and then made it PCO. So, it was like a three-letter company.
Lassman:Was it a moneymaker, though?
No. It was like any startup. It was losing money. The trick was getting up to that pilot production capability from the exploratory development phase, if you will, where scientists can make one or two devices, but now can you make a hundred of them that are all exactly the same. And, we brought in our engineering minions, twenty folks from our engineering department. In the last six months, just put on a full-court press. We had a Nortel contract that, where the price breaks at certain volumes, that Nortel would get on the transmitter and receivers. And, our engineering group was coming down on a cost curve parallel to what we had to do, making great progress. But, by that time, it was just a fait accompli, we had gone too far in the process of shutting the doors on the thing. And then, so, it eventually sold the lines to Nortel. They picked up the technology pieces and continued it themselves. And, my guess is we, we met them later on when we began interactions with Nortel in the late ‘90s. I never asked anybody that.
Well, at any rate, so we closed the doors on that interaction. But, we were still trying to grow the erbium-doped fiber amplifier business, still needed laser diodes. We were out buying them from everybody. And, eventually, oh I guess it was probably a little bit pre-bubble, but we bought laser diode labs. I mentioned we had bid respect to diode labs and bowed out on the bidding at some point. You know, it was just outlandish. But, we did buy a laser diode lab. So, we had a presumed source of laser diodes. But, it turns out all these companies were basically start ups so none of them had a manufacturing mindset or base, and they were making one-offs with very sophisticated, you know, scientific talent on the production line. So, that was a learning experience. I think it’s in some sense, it’s part and parcel of the reason the bubble burst as well. Many bigger items and that things were just overheated in the whole thing. But, certainly a part of it was, indeed, the manufacturing mindset had not, has not, yet, perhaps permeated all aspects of the fiber telecommunication components picture. But the fiber we very definitely knew about learning curves, and we came down a very steep learning curve. And the Duke mantra was that “I want to be half the cost of my nearest competitor.” Cost, manufacturing cost.
Lassman:On just the fiber optics?
Keck:On just the fiber. And that was the mantra that, after Duke retired, I picked up as we were trying to build the photonics business. I said, “You know, that’s got to be our mantra.” And, again, that’s one of the things that didn’t get inculcated in the system before I finally retired.
Lassman:What were the challenges there?
Keck:In growing the photonics?
Lassman:Yes, and trying to negotiate this new environment?
Keck:Well, just size of the organization, and how many times you have to say something. And where you have to say it, and to whom you have to say it, and do they believe it, and how can you convince them that this is the direction things should go. But, it just becomes very much more difficult if you’re a — how big were we? Well, the lab and the bubble was up to 3,000 scientists. When I joined it was about 1,000.
Lassman:So, it tripled in size?
Keck:Well, we doubled the laboratory over a period of three years, as we went through the bubble, as we built up the bubble. Doubling in three years is a pretty good growth rate. And, built new buildings. We didn’t have housing for a plant. And, the operating divisions were doing similar things because we were buying this, and this, and this, and this, and we had to have somebody that was going to watch over them, and make sure things were happening. So, it was just a very much more diffuse organization. So, the problems of communicating these philosophic notions, takes time to, you know, tell everybody once let alone, you know, tell them six or seven times. So, a lot of that was operative. We just grew too big, too fast. And, it’s a dickens of a problem to get everybody singing from the same sheet of music when you’re going through those things. I mean, the whole industry, you know, suffered the same sort of thing.
Lassman:Did you begin to see that there were going to be serious problems given the rapid growth rate in the ‘90s? Just in the fiber optics business, before you become director of research?
Well, no. I’ll be honest with you I was believing my own rhetoric, I think. In my last two, three years of my tenure at Corning, when I was, you know, by this time division vice president, and vice president, most of my time was spent giving “learned technical talks” to Wall Street analysts. I’m using “learned technical talks,” in quotes. I mean these were watered down corporate marketing propaganda things of how Corning was going to win, and what interesting new technologies we had that were going to be more competitive than anyone else in the industry, and it was, it was just high level sales that we were about. And, the thing that, in hindsight you look at it and you say, “Oh gosh. I should have known that.” I remember somewhere back in the line — well, I helped get an organization going, an industry group, OIDA. It’s headquartered in Washington, Optical Industrial Development Association.
I think we formed it in about 1990. It looked, at that juncture, that Japan, Japanese companies were going to run away with the semiconductor laser, transmitter receiver market. And, a group of us got together and tried to figure out, rally around, what U.S. industry could do, or where government could help, and all those sorts of things. And how do we advance the North American optoelectronics industry, photonics industry? Where was I going with that? So, we formed the organization, pointing toward the bubble. And, one point we wanted to pull together a strategy for the industry, and we brought in a couple of MIT professors. One was Charlie Fine, who is a business individual. And, he told us about the bullwhip effect. The bullwhip effect goes along the lines of you’ve got a supply chain. You know, you got the raw material manufacturing, gives the stuff to the component maker, gives the stuff to the [assembly] maker, gives the stuff to the system maker, that gives the components, that sells the components to the [network] maker, and you’ve got this long supply chain. And, the [network] guy knows what the market looks like, because he’s generating it basically. And, he knows whether it’s going up or going down.
Well, by the time he says, through his supplier, “Don’t, ship me ten percent less than you did last month,” and next month, “You know, let’s make it another ten percent down,” well by the time that gets back down through the supply chain, the poor company that’s geared up for whatever the ramp rate had been originally has got, you know, is closing the plant or running it full bore. And, Corning is on the early end of the supply chain, obviously. So, we later, after the bubble burst, I was told that, after I retired, there were various conversations in which the telecom company said, “Well, Corning would promise us, you know, we’d order this much fiber and Corning was having trouble making it. They’d only give us half as much. So, the next order we placed was two times what we thought we needed, because we knew Corning cut it in half.” And all of them were doing the same thing. And they were all expecting they were going to get 100 percent of the pie, and we were supplying several of them. They all said, “You know, here’s our volume, and we’re getting 100 percent of the whole market,” you know. Everybody got a piece of it. By the time that got factored in, you know, we saw this huge demand for fiber. Well, obviously it wasn’t going to come to pass, because each one of them got only a third of the market, or a quarter of the market or whatever it was, and, oh by the way, they had over specified what they needed.
So, the bubble burst, and you know, it burst all the way along the line. And then you find out that there were all sorts of shenanigans going on as to how some of the deals were being financed. But, you know, at the time it was going up, I had a fellow in my group that I had hired from Bell Labs that was charting the growth of the Internet and the number of bits that were flowing, and we thought we had pretty reasonable models as to how much information society was going to need. And therefore, then, they would trickle it down to how many fibers you needed, how many amplifiers, how many transmitters and receivers, and so on. So, we dissected the market, you know, a whole bunch of different ways. And, our models eventually predicted that there would be a slow down, but never, I don’t recall that they ever predicted they would be as dramatic as actually happened.
Lassman:As they were?
Well, by the way, the Internet use is still going up same rate it was. The power users are still on the same curve that they were on, and they’re the lead users on any new technology. And, we had a metric, that when they exceeded it in bandwidth requirement the power users that — that only way you could solve their problem was a “fiber to the where-ever”. The power users have reached it, so then of course, the question is how fast for the average user to approach the curve. But, you know, we got the plots, and so on. So, it looks now as, you know, the market’s going to come back. It’s not going to grow at — what was it growing? One hundred and fifty percent a year. It was absolutely mind boggling the way the thing was going. And we should have known that, you know, the gross national product grows, at what, three percent a year, so anything that’s that many times larger than the gross national product there’s got to wrong with it. Now, naivete was there and we were enjoying the ride. But it was, it was terrific. And when the bubble finally burst and you began to have to let go the scientists that you had worked so hard to hire, and the salaries had risen exponentially.
Normally the dot coms, the start ups, were giving stock to a new employee, rather than an upfront bonus cash payment or something like that. And the industry could usually win out by hard cash, or often, and get personnel you wanted. But, the dot coms were starting to hand out cash bonuses, a yearly salary ahead of time. You know, $100–150,000 dollars. The average salaries they were paying for someone fresh out of school blew my mind. At the end we were paying $100–$110,000 for a fresh scientist that didn’t know anything, just to get them in the door. And two years into the mission, having worked for Corning and learning some stuff, you know dot com would hire them away. “Well, we’ll give you 15,000 shares of stock, and give you a sign-on bonus, and oh by the way the salary will be the same.” So, keeping people during that upswing in the bubble was just frightful.
Lassman:How much did your research budget increase?
Keck:Well, I don’t remember the numbers anymore, to be honest with you, Tom. But, the corporate strategy was we’re going to keep pace no matter what it is. And, the fiber business was, [Lee] fiber was just churning out wonderful profits. So, there was money in the corporation to do that. But, I’d go through my group and I’d have, you know, we’d highlight who the key scientists were. We’d make sure that they were being paid well above the market. We were doing market surveys constantly, it seemed. Certainly every six months. A pool of companies that, you know, shared average salaries and starting salaries and all sorts of stuff. But, for those leaders, we would make salary corrections maybe every six months, give them a boost. And, Corning, to combat the dot com, began using stock options. It used to be that stock options would come down to the vice presidential level and then stop. Before it was said and done, there were technicians that were getting stock options, if they were important to the project and had done a great job. They might get ten, twenty, thirty shares, options, shares. It was an interesting time.
Lassman:You had started out, and it had taken so long for fiber optics to become a viable commercial technology.
Keck:Why didn’t I see it?
Lassman:Well, not, I wouldn’t put it that way. But, it is an interesting contrast given your history at the firm, and then all of a sudden you’re seeing a tremendous expansion which is very different from what your own experience had been.
Keck:And if you plot the curve, it just, there’s this bubble that stands out like a sore thumb. Because now we’ve got data, where the business is growing. Had we kept on a nice steady — I’ve forgotten what we choose — nice steady, pick a number, fifteen percent per year growth rate, which was astronomical.
Lassman:Still it’s huge.
Keck:We’re right on that curve. And here’s this big blip. Yeah, we made a lot of money for a short time, on paper. And then, well we just had a $3 billion write-off.
Lassman:Well, keeping this broad scenario in mind then, let’s move onto vice president and director of research. So now you’re, presumably, you’re not just dealing with the fiber optics business, but you’re looking at the whole technology portfolio of the company?
Keck:Yes. That was fun. And, the trick was, how do we apply the things we learned to all the other businesses, and what, what’s the new look of the corporation. At the time, the bubble was going up, sales were like, they were approaching $8 billion, certainly $7 billion. It had been a $3 billion company. Let’s clarify something here. They didn’t give me the title of director of research. They judged that I was probably already too old. And, what they eventually did was we created a triumvirate, an office of research. So, I think, if I wrote anything down, I probably said research director. So, I was one of the research directors. There were three of us that ran all of the Corning research in the bubble days. Now, having been around Corning as long as I was, my real job there was to mentor these other two young scientists that were working with me. But, we all had the same title. We were all executive director, office of research. So, I don’t know what I said. I generally don’t use director of research, I usually say research director. But, I was a vice president. That’s true, and clearly the senior person on the team, and so I had a lot more sway than any of them did in the directions we were going.
Lassman:When you were director of optics and photonics research who did you report to? Can you carry me through the hierarchy, and how that changed?
I reported to a director of research, that was Gerald Meiling. And he reported to Dave Duke, who was the CTO, and had research, the engineering group, aircraft operations. I mean, you got to put some of these entities someplace that cater to the executive suite. We had a fleet of airplanes down here. Corning is a lousy place to get to as you well know. And so we’ve got I don’t know how many airplanes now. Fewer than we had, but nevertheless we had a aircraft fleet. And they reported to Duke. And there was probably something else. Purchasing. So, he had a number of staff operations reporting to him, and one of which was director of research. So, Meiling was, his title was vice president and director of research. And then, when Duke retired they brought in Skip Deneka who had worked with me in the early days. He was in the manufacturing, helped bring up the manufacturing plant, the fiber plant at Wilmington. He grew up in the fiber business, as did I. I stayed more to the scientific end of things, and Skip, as he went by, was much more aggressive in trying to play the game and move up the corporation, and he did. In Corning, what you found has become the norm is that if you want to escalate your position you’ve got to have a number of different experiences.
Well, basically ideas, walk in somebody’s shoes for a period of time so that you know exactly what that job is about, and another job, and so on, and do as many of those as you can. You know, if you’re anointed as one of the fast movers — at one point, by the way, I had been anointed as the, on a track to be CTO. This was early in my career. But, I chose to stay more with the technology part of the organization, and the scientific community. It was just more comfortable there than playing all the political games, tiger games as sometimes they’re called, politics of the corporation. But, people had thought that I had some ability, some reason to think that you might aspire to CTO. But, Skip had done those things, and had run a business, and done work in the lab, and done work in engineering, and so on. And so, when Duke retired, Skip took over CTO and redid his staff-reporting relationship. Skip liked the idea of teams. You know, the single line just, organizations go through this phase where you have teams and you build consensus as opposed to individuals making decisions, and so on. The theory being that you get a better decision coming out of the team than you do out of an individual, but it takes longer. Well, at any rate, as one of Skip’s…
Lassman:What’s your take on that, by the way?
Keck:Well, it clearly depends on the capabilities of the individual. If you’ve got good people and you trust them, you put them in a position, and yeah they may make a mistake here and there, but I’d op on the side of speed, personally. But, that was not the way that the corporation operated, as I perceived it. Well at any rate, Skip, Skip had a very flat staff, a whole bunch of us reporting to him. I had a group, and lots of others, but then when eventually Skip retired the same time Ackerman did. By the way, I had planned on retiring too, at age sixty, independent of where I got in the corporation. And, they asked me if I would stay on for another year and help bridge the gap, the tide, because here Deneka was retiring, and Ackerman was retiring, and you know, there weren’t any senior leaders that had that much experience in the technology group. They just wanted me to stay on.
Lassman:Ackerman was retiring as CEO?
Lassman:Okay. This is after the bubble?
Keck:This is about the peak of the bubble. They got out right at the peak. Stock was $113. Had I retired at that point, I wouldn’t be living in Big Flats. But, they persuaded me to stay on another year, and I felt that even if it came down, or leveled out it was going to be fine. But, it burst, and within the year the stock was down to, by the time I retired the stock was at $7.
Keck:Yes. So, my retirement looked considerably different, than it would have had stuck to my game plan of retiring when I was age sixty. But, so, at the end my job, we formed this triumvirate of us that were directors of research. Indeed, we operated as a team and ran the entire operation. Was it any different? No. Not really. By that time I knew what was going on in all the other divisions, and for years we’d been meeting as a research team, and a technology team to chart, and decide which were the needed projects and so on. So, we’d been doing this team decision making for a long long time. So, it wasn’t a big change, and my job was to basically bridge the gap with loss of the senior leaders and mentor these two young aspiring scientists. So I did. A year and a half later, I retired. I had a year and a half, maybe a bit longer than that. I’ve forgotten when we created that. We clearly created it before Deneka left. So, maybe two years that I was vice president.
Lassman:Chart for me where you saw the growth markets, at that time?
Well, we were still, to a degree, believing our rhetoric. We had grown from a $3 billion to a $7 billion corporation. And, if you looked at how corporations grow beyond certain points, you basically have to evolve up the value chain. And, material supply companies are the bottom of the value chain, and you got service providers at the top of the value chain. And we were somewhere down at the lower end of the value chain producing not a raw material but a finished component that was then used in other stuff. And so the strategy that the three of us pulled together built largely upon my optics cross — we were still using that as the underlying mantra — and it was that we’d evolve up the value chain to the subassembly arena, to a higher degree. We had already moved the amplifier business that I helped get started in Corning. In the end the laser transmitter business, so that was a bit of a subassembly already. And, we were beginning to assemble a biotechnology presence. One of the fellows that came in, Keith Horn knew that arena fairly well. He was a chemist from Allied Signal.
He began to hire somebody, more biotechnology people, and we began to look at DNA, heredity plates, for possibly some drug testing, optical readout technologies associated with that. Microfluidic DNA arrays is where the reactions are taking place, where there was bonding, and so on, and all the things that the DNA culture was calling for. I saw that as a burgeoning market for all the pharmaceutical, drug testing, biotechnology, medical sensing, arena, and then evolving up the value chain. So, we began too late, in some sense, putting together a research effort and packaging of devices and componentry. We had to bring in new materials technology, begin to assemble the group that in fact would put these things together. And the kick that I was on was, we really needed to get the robotic assembly technology put in place in the photonic arena. Everything was still being put together by hand, basically. I had plotted the learning curve, well I knew what the learning curve was for fiber, and it had just come down steadily — are you familiar with learning curve notion? Back in World War II, it was discovered that as people assembled stuff, and built things, they got better and better at it the more they did it. And, they found that, and they’d come up with innovative ways of doing it faster, as they’d done enough of it. They got tired of it and said, “Well, there’s a better way.” And invented this, that, and the other.
The basic notion is that for every doubling of cumulative volume of a product that you manufacture, you’ll decrease the manufacturing cost by a number of somewhere between fifteen and forty percent. Semiconductor industry comes down an eighty percent learning curve, i.e. it comes down twenty percent for every doubling of cumulative value. Well, that’s performed motion. The problem is that as you make more and more of it, why it takes longer and longer to double the accumulated volume. Of course, if the volume is going up exponentially, why now the time periods remain about the same, and the cost keeps coming down. Fiber was coming down a very very steep learning curve and has continued for thirty-five years to do that. We can probably make fiber now about the same cost as copper wire, which was very cheap. Some copper wires. I plotted the amplifier learning curve. While we came down on some of the early stages, it just kind of leveled out. And we were increasing the volume and planning to go up. We’re making tens of thousands of amplifier units. These were things, oh about the size of the tape recorder. Fiber, and pumps, and eventually had to have all sorts of sophisticated sensors to detect the signal level and monitor the power going in, and all sorts of bells and whistles. But, the costs just weren’t coming down. They flattened out, and the volume was predicted from tens of thousands of units to a quarter of a million amplifiers.
And so, I was putting a full-court press on how do we put programs in place to fix that, and get them, the machine people and the manufacturing people and robotic experts involved in putting it together automatically. Unhappily, the bubble burst before we really got that pulled together. Subsequent to the burst bubble, Corning decided that that part of the business was not likely to be profitable enough for whoever was running it, the leaders of the corporation now. And so, they formed a joint venture with a small company called Avanex. Corning and Alcatel, a huge telecom, largest telecom company in the world, in France, they put in laser diode technology, and we put in the amplifier factory, and Avanex was the small entity that basically bought the technology from both of us, and formed this venture. So Corning — I’ve forgotten what — we owned some fraction of Avanex, and Alcatel owns a little bit larger fraction. And then what had preexisted in Avanex was obviously the kernel from whence it grew. But that part never made any money. The fiber, the amplifier factory, I think, still is breaking even. I just saw the, well it’s in today’s paper, Avanex lost, still losing money. About $22 million this quarter. But, it looked like the market’s turning around and if they’re improving their manufacturing capabilities, which was the tack that pushed them on, then they may recover and become one of the major component suppliers of the industry.
Keck:Corning still has a piece of that. Pre-bubble, or during the bubble, Corning bought out the Siemens' cable venture.
We bought Siemens’ half. The cable part is now totally within Corning, known as Corning cable systems division of Corning, Incorporated. They had built up a capability to do a lot of connector work. They were looking at coupler technology that I invented, helped invent, back in the early ‘80s, mid ‘80s. Their constant — building the ancillary hardware, the boxes that somebody will put in a wall into which you put the stuff. I wouldn’t have thought of that in a research lab, but before they deploy something why you wind up — you know, these pedestals that stick up out of the ground that periodically — look along the highways. You’ll see these little posts, greenish posts. Well those are connector points for a telecom system. It may be amp, optical amplifier in one of those things. Siecor sold the hardware stuff. And basically supplies that market, and as I say a lot of connector stuff and test equipment that they sell. Right now Verizon was wiring up a million homes. It’s been quite widely publicized, with fiber to the home. That’s an overstatement — it’s fiber to the premise or curb, is what they’ve, they’re talking about.
Basically, they’re running these fiber systems along streets and subdivisions so that they pass fairly close to the home, and it’s then a question of running a jumper out to the fiber and bringing it into the house. And then the trick would be, what’s the take rate? Which is basically the way cable television market evolved. Cable television providers, you know, ran along coaxial cable out into the subdivision and said, “Who wants to sign up?” If you’re area like ours where you can’t get, because of the hills you can’t get a television signal from but one station, why, you sign up for cable. And, I don’t know what the take rate on cable television is right now, but it’s probably eighty percent of the homes passed, will pick it up, and subscribe. My guess is that fiber to the home is going to be very much the same, and the cable industry is fighting back. They’re trying to offer telephone service over the cable.
Keck:Voice over IP. And the telephone companies are trying to offer, ultimately, television signals over fiber to the home. Digital transmission, along with your phone service. It’s interesting to watch how these business models evolve and how it’ll all play out.
Lassman:I guess the, in the case of fiber in the home, you would have all, everything coming in over the fiber?
Keck:You certainly could have.
Lassman:And would that have a tremendous benefit to the consumer?
Keck:Yes. My cable modem hooked up to the computer. When I first put it in they, Time Warner, was talking about one and half-megabit capacity, and that assumes nobody else on the cable is using it. If somebody else is using it why mine suddenly slows down. So, you don’t get the full megabit and a half. For the fiber to the home, what are we talking about? I think they’re guaranteeing a megabit and a half to every customer. And the aggregate bandwidth in the fiber is up in the thirty, forty megabits, and then they have a splitter and they’re dividing it among several homes. But never dividing it more than it wouldn’t allow me to get a megabit and a half. Now, a megabit and a half isn’t, may not be all that much. For high definition television that’s going to be dicey. There’ll do a lot of compression. In fact, I’m not sure a megabit and a half will do HDTV.
Lassman:But at least you wouldn’t have the load problem that you have now?
Keck:Not as big as you have on cable.
Keck:Yes. So, where are we?
Lassman:Well, I’m jumping around a little bit. But I want to go back to a comment you made earlier about when you were a research director; this three-person team. You were a mentor? How did you do that? What were the lessons that you brought to these two younger proteges who were coming up?
Keck:Well, that sheet that I gave you on, on my little — we were all close together in the offices, so we had excellent communication among us. We regularly met and brought our staffs in to learn what was going on. We had laboratories all over the world. We had one in Japan, one in St. Petersburg, and England, and France. I had a lab in Somerset, New Jersey. I had one out in California. So, by this time, as the bubble’s going up, the business leaders had acquired all these other ancillary laboratories and we had to integrate those into our thinking. And I, for years I had been, had responsibility for a research group in our France laboratory. So, I knew about global, globalization, and the interaction with outside laboratories, and just could bring that experience base to the group. The information transfer was just interacting, and questions would come up, and I’d give my opinion. If they picked up the things that I was saying, why, you know, they were picking up learning. Everybody knew my — I say most everybody knew my set of mantras. I would repeat. I did that often enough. They’d ask questions about when the fiber business was evolving, “What do we do here?” And so, I had that history to bring to them.
Keck:Just, the normal stuff that someone with say ten more years than the next person brings to the table. It was a very collegial group. We got along very well, everybody listened well to one another, and we were able to interact, and I think we did a very good job of growing the laboratory. The last acts that we did as a group were figuring out, “Okay, we got to cut thirty percent,” “and, where do we cut?” So [we] went through a very thoughtful process and tried to maintain those core strengths that we knew we were going to need, and tried hard not to just focus on the youngest, the newest member of the research staff. We were looking at the skills that we wanted to acquire and where we, what our strategies were telling us we needed to go. So they had an interim CTO, Don McConnell, who had been around for years, and was Skip’s right hand person, if you will. So, he took over for a period of time while the company looked for a new CTO. And then we brought in Joe Miller from DuPont as the CTO. He has that position today. And he seems to be liking it, because the last word I had was that his wife had moved up here. She had some very high level position in the Delaware educational system, the State, State Education Director or something like that. And she’s now moved up here, starting a science, science and math school. And, they’re building a house, and it sounds as though he’s planning on staying at the helm for some period of time yet.
Keck:So, things worked out, ultimately well. The lab scaled back as it needed to, shut down one wing. I was up the other day and they told me that, in fact, they had hired a few people and they were going to open up a few offices in the wing that they shut down. So, as Jamie Houghton came in and got the discipline, needed to get the financial balance sheet back, and operation, and they’ve paid off all the debts that they had. There was a large debt associated with one of the acquisitions that they made. And, as I understand it, it’s all behind them and they’ve got multiple businesses now that are working very good. So, who knows? I’ve hung onto my stock.
Lassman:Well, that’s good. I’m going to move ahead since we’re almost to a quarter of five.
Keck:Not a problem. Let’s try and finish up, but I don’t want to cheat you out of getting questions answered that you want answered.
Lassman:I think if, there are a couple here. Let me, just give me a second here.
Keck:I saw this one, “Importance of university-industry collaborations?” At some point in, I’ve forgotten now what it was…
Lassman:Let’s focus on that for a minute.
Keck:But, I think it was when Meiling was still, was director of research, and there were a group of us reporting to him.
Keck:Dave Duke would have been the CTO.
Lassman:This would have been in?
Keck:Late ‘80s, early ‘90s. I don’t even remember when Duke retired. Isn’t that interesting? Well, I guess I do have a measure of successfully arguing a position vis-a-vis universities, because I always allocated some amount of my budget, directorate budget to spend outside at universities.
Lassman:As research director, or as director of photonics?
Keck:It was photonics.
Keck:I argued that as a research group, we should increase that, double it basically. I wanted to get it up even higher than that. What that would have meant was that as we grew, you’d hire fewer internal scientists and use more of the money outside at some university, and get some professor who was doing work in the field to do work on your behalf, effectively, and of a basic nature. We did increase the amount that we were going to spend at universities. We were doing it as a percentage of the research budget, and I can’t tell you the percentage now. We began to consciously get it allocated as a percent of the research budget. I continued from that point on, even when the three of us were running things the last couple or three years.
Lassman:You mentioned — I don’t know if you mentioned this before when we were discussing this briefly, but was there a specific model that you were looking at, when you were thinking of university-industry collaborations, was there another firm or a consortium type thing that helped, informed your thinking about how Corning should do R & D?
Well, yes and no. Did I hear some other industrial scientist suggesting that this was the direction to go? Not really. It was just, seemed to me that this was, had to be the direction that we had to move the company, that the industrial companies in the U.S. had to move. Well maybe I was picking it up with my interactions with Japanese colleagues. Clearly, in Japan, the industrial work is much more of an applied or developmental nature, and they revere their university collaborators, and defer to them. In many cases, the university researcher, lead professor of the institution, is directing, pointing the directions for his industrial company to go. They had far more clout. I clearly wasn’t on that page. I was, you know, so you saw that happening, and it was just that there was so much to the applied arena that I didn’t see we had enough resources to spend our time doing the basic, as well as the applied, and was poor toward development and so on.
I wanted to make sure we do the things that only industry could do. The other thing that we did along this line, when we formed this OIDA, one of the things that we pushed to put in place with DARPA — particularly in the photonic arena, because the leader of the programs there had some out of photonics, and was a good friend of the fellow who was executive director of OIDA — but, we stipulated that DARPA should insist that the university work had connections with industry. What got put in place as a result of that, was that all of the, DARPA created a bunch of optoelectronics centers, $30 million to six centers sort of — they were a consortia — the stipulations were universities had to collaborate with another university. They had to have a tie with industry. Each one of them set up advisory boards, industrial advisory boards. I sat on two of them. We’d meet semi-annually, I think, for most of them. I don’t think I had any more quarter. We’d go down and listen to the research that was going on, give advice to the university folks and then come back. We had the written reports of what was going on that you could circulate around, and try and make sure that the scientists that were doing things associated with that got out there and talked with them, picked up the technology. That was certainly a move in this direction; let universities do the more basic stuff and have a good connection, then, with industry to be able to pick it up and move it forward.
Lassman:Okay. Just one other question related to this. Infotonics?
Lassman:Could you comment a little bit about your involvement in that, and what that is.
Keck:Well, it sprang out of OIDA, as I describe the history. Some of the other people at Infotonics maybe have a different view, but…
Lassman:And, OIDA, again?
Keck:Optoelectronic Industry Development Association.
Lassman:Okay. Just want to make sure I got that.
So, a little over ten years old. Arpad Berg was a, had evolved to a senior position, staff position, at Bellcore, and in the early ‘90s it looked as though Japan, Inc., was going to run away with the semiconductor laser market. Arpad called a group of us together, down at Bellcore, where he had grown up with in the industry. “What should we do about it?” He had a proposal that we start an industry association to band together and figure out if there weren’t some common things that we could push. The model being Sematech. When the semiconductor industry was about to go under, they all got together, agreed to pony up with lots of money, and the government matched it, and so on. We formed this industry association and there was a steering committee that, initially I sat on the steering committee, and we had, I’ve forgotten now who was on the board from Corning. Well, it was one of the business leaders. I don’t remember now whom.
Well, the steering committee, basically, you know, was a working group that got together and posed, “What strategic direction should OIDA be taking?” And, Bob Leheny, who was the DARPA guy that was funding some of the programs, particularly those related to mapping strategy, wondered aloud at one meeting whether or not we shouldn’t have a photonic Sematech, with industry ponying up, and he, DARPA, thought he could come up with $60 million to kick in and be a co-funder, and couldn’t we put together an entity that would become a research arm for the photonics industry, ala Sematech, that all industry could make use of? That idea lay dormant for ten years, eight years. Then, at one point in the evolution of the organization the board share was disappearing, and they asked me if I’d take over the chairmanship of the OIDA board. I’d already given up the steering committee, but was committed to the organization, the OIDA, and funded it out of my budget, with annual dues.
When I took over as board chair, it struck me that we were still in the same boat, in many respects of, we needed some entity in this to do work of common interest to the industry. I pushed on that, at the board level, and I had the luxury of having one of my managers on the steering committee now. It turned out there was a fellow from Kodak, David Smith, who was also on the board, and he had one of his managers, a Washington lobbyist actually — well, not a lobbyist, but a Kodak person that kind of enables introductions with the right agencies to get funds and submit proposals, and whatever. Rick Jarman was David’s appointee to the steering committee. We got to talking, and basically Tom Holmes, my guy, and Rick Jarman began to flesh out what this photonic Sematech thing should look like. At the time, switching technology was important. Corning was interested in switching, Kodak was interested in display technology and it looked to us as though this MEMS technology -– micro-electromechanical mirror technology –- silicon micro-mirror technology was the way to go — you don’t look as though you know about this technology.
Lassman:I’ve heard, I’ve heard of it, but admittedly I’m not fully up to speed.
Well, at any rate, it’s, it turns out you can make little micro mirrors using silicon wafers, and move them and tilt them and focus them and all sorts of things, and they’re display devices. They’re just on a —. So Kodak was interested in display. We were interested in switching. And “Oh gee, let’s see if we can’t push this forward.” And David was the one I certainly give credit to for saying, “Well gee, now Rochester used to be known as the capital of optics,” “And, it’s not anymore.” And that was one of the other sad realizations that Arpad Berg, executive director of OIDA had brought to him, to David and I, and Arpad’s conversations Rochester was not known as much of anything in optics anymore. Well David said, “Yeah. We got to change that.” And, “Let’s locate this thing in Rochester.” About this, and I don’t know how it came about. Rick Jarman, as I say, had all sorts of connections on Capitol Hill. The next thing I knew Chuck Schumer, senator from New York State, was pawing around, came into Corning, talked to Ackerman, and his successor John Luce. Well, at any rate, I don’t know whether Schumer approached the Kodak executives. He may well have, but at any rate, suddenly we now had a receptive federal official. David and I got to talking. He knew a fellow at Xerox, Steve Bolte, who is now director of research at Xerox. The three of us got together and we scoped out what this thing would look like, and came up with a number and it was going to be about $300 million of investment that would be required. And, the notion — this was still as the bubble’s going up. So, you know, the telecom bubble is going up.
Kodak’s already suffering, starting to suffer. Xerox has suffered, and has hit the bottom. But at any rate, so $300 million was what we scoped it out at, and I said, now the formula we’ve got to use is we need a quarter from state matching, and get the state to match industry, and then we’ll get that, take that pot, and go down and get the federal government to match that. So, $75 million each and $150 from the feds, and we got our $300 million. So, that was the strategy we began to roll out. And, things moved — this was 2000, so the bubble is still on its way up. Corning looked like we could spin off $15 million in petty cash. I imagined that where we were going to get five, five New York companies. We need the three of us plus GE and IBM. And, we began to get the $75 million, and the technology was going to be silicon-based processing to make MIMS devices. And, in the back of my mind, it was going to evolve to the point where we would bring in gallium arsenide and indium phosphide technology to do development in laser, for lasers to transfer to our Lasertron facility. And Kodak could see the reasons for doing it. Xerox was rather mum about exactly what they wanted, but inkjet printing was high on their hit parade. So at any rate, that was the thing we drew up.
And, about this time, we got wind of the fact that Governor Pataki was talking about a large sum of money that we need to fund universities and elevate them to some new levels with new technology centers, centers of excellence. We wrote state officials, a fellow at Corning, a state lobbyist, knew people, and so we sent them to Albany and began to put this idea forward to the Governor and low and behold he put out a, in his State of the State address he basically called out for the photonics center and said this is one of the ones I’m going to fund. And, we already had our commitments. Now we had to sign, we had to have the CEOs sign before January 2001, when Pataki gave his address. And then we began to petition Schumer, and Clinton, and Congress to get the earmarks coming our way. The first year we had $20 million in earmarks from the feds. We were assuming $30 million for five years, to get the $150 [million]. Unhappily that’s come more slowly than we wanted it, and by this time the bubble burst, and Pataki didn’t have any money. And, Xerox had a building they wanted, they got out of the inkjet printer business and wanted to give up the building, write it off.
And, it turned out it was exactly the footprint of the building we were going to build for $60 million, and they wanted to sell it to us for three. And then take the rest in write-off. So, in fact, that’s what we did. But, it took us a year to get the money from Pataki out of his slush fund, because 9/11 has happened by now, and the whole New York economy’s in the tank. Corning was going to kick in $15 million in cash, which we desperately needed. The bubble burst and the cash disappeared. Kodak and Xerox had always been planning on doing in-kind donations to get this thing off the ground. And, that’s, they followed through with that. Corning followed through. A lot of our photonic equipment that, when we downsized the photonics activity, went into the Infotonics center. We donated that. In some cases we had equipment where the boxes hadn’t even been opened. You know, brand new stuff. So, Infotonics benefited in that respect. I had a cockamamie notion that we should accept anything that industry wanted to give to the Infotonics Center, almost, and set up a lending system where a university, if it needed a spectrum analyzer, an OTDI, or whatever, could come up and borrow one of these things. And so we wanted to have a strong linkage between Infotonics and industry, and universities to bring in the ideas, and this was going to be a seed bed where these ideas would come in, industry would look over it, and say, “Yeah, I like this one. I like this one,” and pick up some of them, begin to do joint work. And so it’s again building on this notion of collaboration, but forming the vehicle whereby you could actually do some productive research, and outside of, outside of. And basically, the main reason was to share expensive equipment. MIMS technology was very expensive to maintain, and all three companies wanted it, and why all three invest. Why not just share the equipment, and get time on it.
So, we kicked it off with the State of the State address in January 2001. That’s probably the genesis — well, yeah, by June we had filed our incorporation papers and non-profit status papers of 2001. Today it’s — and Duncan Moore was the original CEO, and when I retired in ’02, I came in as CTO, part-time, actually signed on as a consultant, but they gave me that title, and we funded $5 million in university grants around to get the ball rolling and come up with some viable projects while we got the capital out of earmarks and company equipment to build the facility, you know, get the silicon foundry going. And, they turned — I stayed with it until July of this year. David Smith retired from Kodak. He took over as the CEO position. Duncan went back to the University of Rochester. And, I bowed out as CTO at that time, and told them to hire some young people to actually do some work. And, they, as far as I know, they’ve begun turning out some wafers. We’ve bid a number of government projects, from Homeland Security to energy, Department of Energy, NIH. We began trying to find other partners.
We had approached GE and IBM, and both of them had, well IBM had invested heavily in the Albany semiconductor facility. You mentioned these big wafers? Well, putting in a twelve-inch wafer facility at the University of Albany, in the center, high-tech center that one of the governors, fraternity buddies, started at the University of Albany. So, they got money, and it’s also in Joe Bruno’s district, who is the Speaker of the Senate, leader of the Senate, Senate Majority Leader in Albany. So, they got lots of money. But, and so far I haven’t gotten IBM or GE in Infotonics. But, we still have contacts and talk about it. So. Well, I think it’s going, and obviously, I hope it succeeds. This is, this whole area is economically deprived. I guess we were, the industry’s reason for doing this, in addition to sharing the equipment, was we just needed high-tech equipment that graduate students from universities could work on, and if we can attract them to the Institute of Optics or MIT, or Syracuse, or Cornell, to work and possibly come up and spend time using this equipment, state of the art equipment, there’s a high likelihood that some number of them will want to stay in the area. And they become now the pool of employees for the companies. And, you know, at the time the bubble was going up, we were, you know, I used to liken it to the water sloshing in the bathtub. We’d hire a bunch of folk and pay them exorbitant salaries and the next thing we knew JDS or Lucent would offer a higher salary and some number of them would go over there. And we were just moving the same pool of people, group of people around. I mean the universities were just not turning out enough to take care of a 3X increase in the industrial size the bubble went up.
Lassman:And is part of that also to encourage these, new young researchers who would come here to study and also to establish small firms, and to build up that?
Keck:Yes. At one point, I talked about Silicon Valley. We — it was facetious there — a number of organizations in Rochester bent on economic development as Kodak and Xerox have gone up and down, and up and down. So there’s was always a pool of employees that have lost their jobs that want to start small firms. So, I had, in obvious mockery of Silicon Valley, I was going to have Silicon Valley and this was going to be the optics corridor, from Rochester to Corning, with Cornell University and the Institute of Optics, and Alfred’s University just, you know, about on the line. But, that name didn’t catch on.
Lassman:Well, it might.
Keck:Well, I’m not around to push it anymore. So, I doubt it. But, what we finally came up with, and it was interesting, as I said Corning had lots of money at the time the bubble, and Bob Ecklin, who became the anointed corporate board person on the Infotonic Center, was, well hired a New York firm to come up with a name for the center. And, so we went down to New York one day and met with these ad agency people, and told them all about our technology and when they came back with a presentation of how you might name it, according to the geography and according to technology, and according to this, that, and the other thing. They had a whole bunch of names for us. Not one of which was the Infotonic Center. But, they did come up with the notion that it’s, was something on the “technological throughway,” the New York Throughway, the toll road that runs from New York City to Buffalo. And, we’re right on that. And that, I’ve used that in a couple of talks. The Infotonic Technology Center on Technology Throughway. And so, who knows.
Lassman:Or the Information Throughway?
Keck:Well, they could have come up with that.
Lassman:Well, do you still have some connection with it?
Keck:Well, I still have a badge. I can still walk in the door. The hour and half drive was getting to me. An hour-and-a-half each way. I don’t envy your commute to College Park.
Lassman:At least I don’t have to drive. That’s the advantage.
But, when David came on board, my guess is that he recognized what I had recognized also, that we were going to be cash-starved on the operation. I mean we spent a goodly amount of money on the capital investment. It’s in now, and behind us, and stuff is presumably operating. But, you know, he really needed to conserve his cash while he gets to see how many of these contracts we’re going to land, and builds up the contract base. So, you know, there’s another activity going where I’ve brought ideas to them. Well, this Mike Barnoski that I taught, you know, that I formed PCO with, has another idea, and might be able to use Infotonics, so I put them together, and had a meeting over the summer. So, the answer is I’m still very much spiritually connected to it. And, if they see a spot where I can help them, I’m happy to help out. But, my guess is he probably wants to conserve the cash a little bit, and spread it out. That’s my guess. But, hopefully he’ll succeed, and once again get Rochester revved up as the photonics capital of the world.
We always had the notion that this was going to be a national center, but knew it was going to take years and a lot of hard work to establish that sort of premium position. There are other centers that are vying for the same sort of thing. So, Christina Johnson is dean of engineering at Duke, and one of the dot com benefactors, beneficiaries I should say, became a benefactor of Duke and gave them, I forget whether it’s fifteen or thirty million to create a photonics center at Duke. And, did the same, gave the same amount of money to Stanford. MIT had notions of a center, so at one point as we were beginning to pull ours together it turned out there were all sorts of other centers that were popping up with the same basic notion, economic development in a particular area. And, frankly, you look at the tealeaves right now, and the nation’s going to need that, desperately. You know, we’ve got the jobless statistics that we’ve got people out of work. The population’s growing. The number of firms that are being created is not growing.
The visa policy that we’ve got, the immigration policy we’ve got, is stopping students from coming here to get their education and contribute their knowledge, and so on, to the nation, and those that have come over, did get through the visa system, a higher percentage, like thirty percent of them, are going back to their home country and taking the technology with them, whereas they used to live here and start the firms, and grow the economy. I don’t [know] where we’re gonna — I’m fearful, you know, that we’re going to have all the boats rise, as the global economy progresses, rather than, you know, one go up and one go down. Because if we keep out inventing and developing new ideas and the new technology, and have these specialized jobs and the expertise to run them — and right now we’re in a sort of usurp the country. When that happens…
Lassman:Evaluating your career at Corning, what you think your greatest accomplishments have been, and any regrets, lessons learned?
Well, I guess — it’s hard to say. It’s hard to talk about your own accomplishments. The thing I felt very good about, one of the fellows who spoke at my retirement made the comment that “A few people in Corning have started a business, a monumental business. Rarely has anyone started two of them.” And, obviously, the fiber business and the photonics are what he was attributing to my efforts, and obviously lots and lots of others. But, I had enunciated, done the technology work to go on, and then enunciated the direction, and the vision for the second one. So, in looking back on when I came out of academe and knew I wanted to do something of value and contribute somehow, make use of the knowledge that I’d learned, clearly never had any vision that you’d something that might actually revolutionize something. And, you’re looking at, hey do a good scientific work and hopefully it materialized and was something that people would use. Just the luck of the draw. I mean, of all the companies I could have picked why did I come to Corning? Why at that juncture in time? I did some good work, hard work, lucked out, made this fiber, landed with Bob Maurer, Pete Schulz, each of whom brought, you know, their important skills to the problem.
The, room-temperature semiconductor laser diode was demonstrated within three months of us demonstrating low-loss fiber. Now, I’ve often thought back — you know, I told you about these helical microwave guides. The electrical guys were coming up with their own solution, and they were doing field tests already. I mean, suppose we had invented the fiber and the laser diode wasn’t around, well, we wouldn’t have gone anywhere. Gas lasers weren’t going to be practical. We invent the semiconductor laser diode without a transport, and it probably wouldn’t have gone anywhere. Two of them together. And then history has later shown that the Internet was demonstrated in 1969, right, a fledgling experiment? Carlotta who said that Intel was formed in 1971. So, here you got this juxtaposition of four important pieces of technology all having their origin within a year of one another. What is the probability of that happening in history? That you just happened to be there? So, clearly there’s an awful lot of luck that’s involved. What is it in nature that allowed the same emitting wavelengths of semiconductor laser diodes to be able to be matched to the point at which the fiber has its lowest loss, and why was nature kind enough to allow the zero dispersion point to occur at exactly the lowest loss. It keeps bringing me back to this notion that Feynman espoused in the messenger lectures that I think I told you about.
This wheel in physics and theology were juxtaposed and what physics couldn’t explain theology had to. But when you sit there and you ponder, oh boy. You know. So, I feel, obviously wonderful about the revolution that we’ve participated in, and the fact that there were two businesses that Corning ran and made a ton of money on, that you had some hand in. I would like to think that there were a number of process things that I put in place at Corning. I don’t know whether they’re still using it, but one of my bets all along and I suspect it’s on the list of quotes you have, “Nothing moves faster than our ability to measure.” “Nothing progresses faster than our ability to measure it.” Obvious, truism, but I tried to, even as a research manager tried to come up with ways in which I could measure the progress of a project. And show whether we were actually making progress or not. One of my managers actually came up with a cute thing based on a thing that sprang out of my work on fiber, actually, our work in fiber. I told you that we charted, as a function of time, the loss rate. Clearly we had, in the early days we had a focus on attenuation rate of the fiber. Then the next thing is, well it’s the bandwidth. And so we began measuring dispersion, and what’s the strength? The project always had some piece of technology where we would, there was a defined variable to measure.
We put the programs in place to measure that variable, and we identified what the important ones were. What we found was that, in looking back at the history of a whole bunch of projects in the department was that the successful ones always had that measure, and once they achieved an element of success on the measure they added another variable, and began to measure that, and then add another one, and measured that. We began to plot a scorecard on a project, and we also said, “You better look at your competitor. Who are your nearest two competitors? Academic or industrial?” I didn’t care. Government labs. Tell me who they are, read the literature, and tell me who your competitors are. And, the simple scoring was, you get a point if you’re ahead of your competitors. You get a point if you’ve hit the target. And, for every variable, you get to add the points for the same sets of scores. And what we found was that after three years, without question, by simply tabulating that cumulative score, you could easily tell the projects that were going to succeed. Other projects, they might, one year they’d be ahead, and the next year they’d, ahead of their competitor and get a point, and the next year they’d be behind and they’d get minus a point.
They’re the same as the competitor at zero. And, head of the target or behind the target, and so on. And, the losing projects, you know, just one year it’d be up, and the next year down. And, it would just oscillate around. But this, the steady performing ones, and by adding more and more variables you got a higher and higher score. And your score as well. So, in two years you could begin to speculate that this one’s going to make it, and this one, these others aren’t. In three years, it was just unquestionable. I tried to get that inculcated in the system. I have to tell you the basic researchers just fought that; people that came from the basic research end of things, just fought it tooth and nail. “You can’t program invention.” I don’t buy that. I think you can, if you’re smart about the question that you ask. “What is it that I’m trying to prove or find?” If you spend the time developing what the hypothesis is that you’re trying to prove. My argument, then on their project was, “Well, if you can’t define what it is you’re trying to accomplish, how do you know whether you’ve made any progress?” You know what the basic researchers are saying, “You can define a measure. Even if it’s nothing more than saying I’m going to investigate three, three new glass compositions and tabulate the results.” “Okay. Well, is your glass better than somebody else? Clarity? Index?” “I don’t know.” “Well, tell me what it is.” But, I had the darndest time ever trying to get that into the largely materials groups, and so on. The physics guys, physicists are grown to measure. They picked up on it. Like I say, I had one manager that would, really believed in it, and had carried it on. But, he retired six months before I did, so whether they still remember that or not — and we developed all sorts of other, you know, stoplight charts to indicate progress. I was at a Red Cross meeting the other day and one of the people used to be in the group, was still using the stoplight chart-to-chart progress on things. For the Red Cross. Perhaps that still caught.
We truly tried to put some good process stuff in place, assist research, and not get in the way, but still try and focus people and get them to think about what it was they were actually trying to do. As opposed to just, you know, measuring something willy-nilly. Regrets perhaps are that not enough of that got inculcated in the system, than it did. But, I guess the, any other regret, at one point I was bent on a third market area and that was this whole sensors area, whether biotech or physical sensors, or what have you. It seemed to me that one of the attributes of fibers that, the ability to transmit signals over long distances — well the civilian infrastructure. I was envisioning you could use fibers and do all sorts of interesting sensing in the civilian infrastructure. At one point I had the notion, I tried to get my daughter to pick this up as a PhD thesis at Cornell. She was a civil engineer. I think we mentioned that at lunch the other day. Her professor up at Cornell wanted her to continue, would have funded her to do a PhD, and was interested in fiber optics. And when Mary Simpson found out that Lynn’s father was involved in fiber optics, why she was even more interested. And the whole notion was, that I was pushing, was could you put fibers, imbed fibers in concreted bridges, and over time test the stress, the strain level in a bridge, and ascertain when it was going to fail? Use that as a real-time monitor.
So there are all sorts of questions that had to be answered. Well, how do you bond it with the concrete? And could you measure it? And what technique would you use? And so on. But, we had developed a whole bunch of biograding, writing capability in the fiber. And OTDLs, optical tunable delay line reflectometery. We could send signals in, and look at return signals, and that could be, with gradings, that could be wavelength dependent, and if you stretched the grading or changed the wavelength that that reflected. Well, it was lots and lots of work in the field that would suggest that there were devices, systems that you could build that could help other areas. Pollution chemically monitoring. I envision that fiber optics fluroimmuno-assay thing. Could have pulled some variant on that theme and monitored landfills were you getting seepage out of the landfill? Well, just a host of opportunities. I participated in an evaluation panel that the government, a number of government agencies funded, the so-called J-Tech Panels. I don’t know if you’ve seen those. George Mason University runs them under contract to whatever agency wants the study run. They basically evaluate different pieces of technology and look at where the U.S. technology is compared to that around the world. And so we looked at [it], in the optoelectronics arena, and they asked me if I’d join the panel. I did one on basically sensors. This was what the agencies wanted. So, we toured Japan.
In this case, it was evaluation of Japanese technology versus here. And I had built up a whole story that I brought back to Corning on opportunities in the fiber sensor area. But I never ever could get that to catch hold in part of the business. I had it caught at one point. We bought the small company in Woodinville, Washington. It had a piece of fiber sensor technology. Les Gunderson went down to the division, same guy that helped me start the laser diode lab PCO bought this small sensor company. I went up many times, visited them, and the hope was that was the springboard where I would be able to push more of my sensor technology. But, unhappily they didn’t, just had trouble making the stuff. They had a neat idea for a sensor, but couldn’t make it. Eventually, he pulled the plug. So, that’s a regret, I guess, the biggest one that I see. That and not retiring at age sixty.
Lassman:If there is anything else you’d like to say, or talk about, I’ll leave that open.
Keck:Oh, Tom, at this point I think my voice is getting kind of hoarse. So, I probably ought to stop.
Keck:And as I say, I probably may have to excise a lot of these stories, and so on. So, I apologize a bit for your, for what we may have to go through.
Lassman:Well, that’s fine.
Keck:I need to — Corning. It’s a wonderful company. I mean, I marveled at the end of the career, and as we went through the 150th celebration of the company, how many companies in the world are 150? Have been at it for 150 years?
Keck:Average lifetime of a Fortune 500 company is thirty years. Before it’s bought, or goes away or, you know, whatever. And here’s one entity that’s been around for 150 years. How do they do it? I saw the slide in here earlier that came out of the Economist article. Where is it? Here it is. Here’s the slide. So, these were the waves of technological advance that the Economist had put forward, somewhere in ’85, over at Palmer Textiles. Corning was formed in 1851. And, steam and rail and steel were coming out. And what did Corning do? Well, opportunistically we took their material and began to apply it to the railroad.
Lassman: Railroad industry. Right.
Keck:Right. And so, they rode that wave of borosilicate glasses, the forerunner of Pyrex, and built the company. They began the laboratory in 1905. And, what’s the upswing here? Electricity, chemicals, internal combustion engine. Well, in chemicals, they began the low silicate, the Pyrex glass, the glassware.
Lassman:Cookware. Things like that.
Keck:So that came up and loosely said it was associated with the upswing.
Lassman:Oh, and light bulb blanks. The ribbon machine.
Keck:Yeah. The ribbon machine was in here. Electricity, electrification.
Keck:In twenty. So, we were opportunistically rode that wave.
Lassman:And tubes? Right? Tube blanks, for vacuum tubes?
Keck:Well, the electronics is the fourth wave.
Lassman:Yeah. I guess that’s…
Keck:Probably, you know, perhaps the seeds of the technology were growing here, but I suspect the vacuum tubes more in 1950s time frame with the advent of electronics. Well, and with electronics, of course, television.
Lassman:The television business.
Keck:And up came the television. And, then you go, grow out of this — she says 1990. In the book, she had it 1971, that she started in the information age and so on. And then of course, the we  with the book. This was the Economist that recorded that. But it was based loosely upon her stuff. Oh, no, this is Schumpeter’s.
Lassman:Oh, Schumpeter. That’s right.
Keck:And she quotes him. . Okay. I didn’t notice that until I read her book. She departs a little bit from, showing the waves getting closer and closer together as technology gets more advanced. She had a fairly uniformly spaced, as I recall.
Lassman:Creative destruction. That’s his — Schumpeter is…
Keck:Yes. His thesis.
Lassman:So, yeah, Corning…
Keck:And here is…
Lassman:You could write Corning history right here, basically, and chart it.
Keck:Well, I gave a talk at the business school up at Cornell, and according to history as posed by Schumpeter, with Corning businesses. I had illustrations of the various businesses. That’s how you do it. You out invent yourself. So, to, I mean, of all the companies to whom I had gone. Martin Marietta’s gone. Lawrence Livermore, well I don’t know, it’s still around, but it’s not known for anything in particular perhaps.
Lassman:IBM is still…
Keck:Still around. It’s gone ups and downs, and so on.
Keck:Goodyear — their lab was decimated. I don’t think they’re doing anything in the aerospace anymore. The CIA is still around. Oh Lord, I wouldn’t have had anywhere near the fun and the exposure. And, for a young scientist, you know, the first job you take in industry, to have it succeed, and succeed to the degree that it has, my Lord. You look back and it boggles your mind.
Lassman:Well, by today’s standard, too, most people don’t stay at the same business anymore?
Keck:Well, that’s right.
Lassman:You know, that’s a generation difference also.
Keck:Very true. Yeah.
Lassman:You know, you don’t you know, your career is very unique in that, I mean by today’s standards, is unique. You don’t think about staying at the same place.
Keck:Although, you will still find people that come to Corning, are wedded to Corning, and stayed on. Something about the small town, and the atmosphere. And then the trick then is to keep then energized, and continue to get lucky outside. As Satchel Paige says, “To see who may be gaining on us.”
Lassman:Yeah. That’s right. Well, I think that…
Keck:A marvelous company.
I think that…
Session I | Session II