Oral-History:Tatsuo Izawa

About Tatsuo Izawa

Tatsuo Izawa is a pioneer in the field of optical devices and components for communication systems. He invented the electro-migration method for the fabrication of planar lightwave circuits and the Vapor-phase Axial Deposition (VAD) method of optical fiber fabrication. He also conducted pioneering work on silica-based planar lightwave circuits. Izawa has spent the majority of his career as a researcher and manager at the Nippon Telegraph and Telephone Corporation (NTT).

This interview begins with a brief overview of Izawa’s education, which culminated in a Ph.D. in Electronics from the University of Tokyo in 1970. Izawa discusses his doctoral thesis on infrared holographic memory and his early research on semiconductor laser dynamics and optical fibers at NTT. He outlines his tenures at Musashino and the University of California at Berkeley, and his return to NTT. Izawa discusses his work in the development of the modified chemical vapor deposition (MCVD) and outside vapor phase-deposition (OVD) methods. He also details his invention of VAD and compares his projects with Corning’s ATT method. The interview concludes with a discussion of Izawa’s current work at NTT, including his current roles in research and management.

About the Interview

TATSUO IZAWA: An Interview Conducted by William Aspray, Center for the History of Electrical Engineering, May 17, 1994.

Interview #200 for the Center for the History of Electrical Engineering, IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc. and Rutgers, the State University of New Jersey.

Copyright Statement

This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.

Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, 39 Union Street, New Brunswick, NJ 08901-8538 USA. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.

It is recommended that this oral history be cited as follows:

Tatsuo Izawa, an oral history conducted in 1994 by William Aspray, Center for the History of Electrical Engineering.

Interview

INTERVIEWEE: Tatsuo Izawa

INTERVIEWER: William Aspray

DATE: May 17th, 1994

PLACE: NTT Opto-electronics Laboratories, Tokaimura, Ibaragi Prefecture

Family & Education

Aspray:

Would you begin by telling me when and where you were born, and what your parents did for a living?

Izawa:

I was born in 1941, December 24th, in Tokyo, just after World War II started.

Aspray:

Yes.

Izawa:

So the situation was very bad. My father was a designer of lighting for buildings. My mother was a primary school teacher.

Aspray:

Did you have brothers and sisters?

Izawa:

Yes, I have two brothers. My elder brother is a high school teacher in earth science. My second brother is an engineer in electronics, and now he is working at Seiko. Seiko is a watchmaker. My daughter is a medical doctor. She is a graduate student in physiology. My son is a businessman, working in a bank.

Aspray:

Can you tell me about your early education?

Izawa:

Yes. I entered primary school at Togakushi village. It's in the countryside in Nagano Prefecture. When I was born I was living in Tokyo, but when I was growing up the war was severe, and we moved to Nagano prefecture.

Aspray:

To be out of the way of danger?

Izawa:

Yes. But after two months of primary school, I changed to a school in Tokyo.

Aspray:

I see.

Izawa:

In 1954, I entered junior high school in Tokyo, and in 1957 I entered the senior high school in Tokyo, called the Nishi metropolitan senior high school. In 1961 I entered the University of Tokyo, and after that I entered graduate school, also in the University of Tokyo.

Aspray:

Were you a good student at school?

Izawa:

Okay, yes. But when I was young, I was not so good in health. I had asthma. I could only study. I couldn't play with my friends.

Aspray:

Did you build things? Did you have science hobbies? Engineering hobbies?

Izawa:

Yes. My brother also had hobbies of engineering, like making radios, or small motors winding wires. At first I watched him, but after a while I did similar things.

Aspray:

As you were coming through school, did you have a particular interest in mathematics and science subjects?

Izawa:

Yes. I think I was a very good student at mathematics and physics. But maybe a bad student for the Japanese language.

Aspray:

The study of writing?

Izawa:

Yes. I was very poor. I am poor, I should say.

Aspray:

So was I! How did you choose to go to the University of Tokyo?

Izawa:

It's not so easy to answer that. Many friends around me entered Nishi senior high school. It was a very famous school to enter a famous university, such as Tokyo University.

Aspray:

I see, so it was a school that was competitive to enter and it trained its students very well.

Izawa:

Yes, that's right.

Aspray:

Many of those students went to the top universities.

Izawa:

Yes. But at my first trial of the entrance exam I failed because at the day of the exam I got a severe attack of asthma. For one year I was taking care of my health. That was my job at that moment.

Aspray:

When you entered the university were you planning on studying to be an engineer?

Izawa:

Oh, not so easy to decide. I was thinking I should go into chemical engineering or the physics department. I liked physics and electronics and chemistry.

Aspray:

Did you have to decide before you entered the university what subject you were going to be in, or did you have some time after you became a student to make those decisions?

Izawa:

In the case of the University of Tokyo, maybe we spent one and a half years on general education courses. We took a general education course, I should say. After that we decided the course, and it depended on the grade of the exam.

Aspray:

It's not only your choice, but you also have to do well in the examination?

Izawa:

Yes, there's some competition.

Aspray:

And you studied electronics, is that right?

Izawa:

Yes.

Aspray:

What did a curriculum in electronics include at that time?

Izawa:

Electromagnetic things.

Aspray:

So the basic theory of electricity and magnetism?

Izawa:

Yes, that's right. And circuit theory.

Aspray:

Was there a strong emphasis on telecommunications? Were there computers in the curriculum yet, as part of the electronics?

Izawa:

At that time the computer was not so popular, and communications is a very special field in electronics, so it was not emphasized. The reason I joined NTT (The Nippon Telephone and Telegraph Corporation) is that the NTT research lab has very big money for research. That's the only reason!

Aspray:

I see.

Izawa:

So I didn't care about communications actually at that time.

Aspray:

I understand that you stayed on at the university for your masters and Ph.D. degrees. Can you tell me about the work that you did for those degrees?

Izawa:

When I finished undergraduate courses it was 1965. In the 1960s, the laser was first developed, and the red ruby laser was first demonstrated. I knew that, and I was very interested in the laser. I broke a red balloon with a ruby laser, and it was very interesting. So in the graduate course I decided to do research on lasers. At that time, the helium neon laser was not so popular, so I built this laser by myself from glassworks. It's not easy to align the mirrors. Very difficult. I tried many times.

Aspray:

I take it that your orientation is towards being an experimentalist rather than a theoretician.

Izawa:

Yes. I don't like many equations. Like the Schroedinger equation. I couldn't understand it.

Aspray:

My technical training is in mathematics. I could never get anything to work in the laboratories. What did you have to do, build something, or write a dissertation to be awarded the PhD at the university?

Izawa:

Yes. A thesis

Aspray:

What was your topic?

Izawa:

Infrared holographic memory. I developed some special materials to record infrared light. I made an infrared image, of a three dimensional image. I developed materials just like a photographic film, but they were very special materials. I made an infrared laser. It's not special, but I built it by myself. I wrote papers about the characteristics of the recording material. That was my doctoral thesis. But it was only a small part of my work when I was a graduate student. I also built an argon laser. It's very big. It was not so easy to make, and at that time it was not popular in Japan. Even in the United States. I got a very big cathode and I made a pretty big base.

Aspray:

What was the purpose for building it?

Izawa:

The argon laser can get the blue light. With that I thought we could make a very high-density memory, but it was not so easy.

Aspray:

Is it unusual in Japan to go to the same school for your graduate study as you go to for your undergraduate study?

Izawa:

I think it was very popular at that time.

Aspray:

It's discouraged in the United States.

Izawa:

I know that. Some Japanese universities do not have graduate courses. In that case they change schools.

Nippon Telephone and Telegraph Corporation (NTT)

Aspray:

When you were finishing up with your doctoral degree, what did you intend to do for a career?

Izawa:

I was thinking I should be a university professor or a researcher in a laboratory like NTT or Hitachi. But at that time we had a big struggle between the professors and the students in the university. Just like the hippies at the University of California, Berkeley.

Aspray:

Yes, right. The same thing was going on in Berlin also.

Izawa:

I was disappointed with the professors, and I decided to be a researcher in big laboratories.

Aspray:

Did you have many choices? How does one go about finding a job to do what one wants to do?

Izawa:

I had an older graduate from (my Alma Mater) my professor's laboratory who helped me.

Aspray:

It's a common path that I hear from many people about how they got their jobs.

Izawa:

A guy recommended me to come to the NTT lab, and another guy told me, "Why don't you come to the Hitachi central research lab?" So I thought, "Which one is best?" My decision was, "How big is the research fund?" At that time NTT was a public corporation. It's a semi-governmental organization, so it was very stable. It had very big money for research. A private company like Hitachi also has big money for research, but it's not so stable. So I decided to join NTT.

Aspray:

What was your first position within the company?

Izawa:

Oh, it was as a very lowly researcher.

Semiconductor Laser Dynamics and Optical Fibers

Aspray:

Were you assigned to a particular laboratory, and did you get a particular problem to work on?

Izawa:

Yes. I was assigned to do research on semiconductor laser dynamics. Characteristics of laser behavior. But when I entered NTT the semiconductor laser had just been born. Before that, lasers only worked at a very low temperature, but in that year some researcher in Bell Labs — he's now in Japan, Dr. Hayashi, made a room-temperature operation laser. My boss asked me to research the dynamic characteristics of that laser because it's very good for communication. At that time we couldn't get a nice laser. Some guy was making semiconductor lasers for room-temperature operation, but they failed and failed, so I was waiting. I started new research by myself. It was related to optical fibers.

Aspray:

Was this something that was unusual for someone to do in the company, or did they allow people to do this? Did they encourage people to start these independent projects?

Izawa:

Maybe it was allowed, but usually the researcher in this lab was assigned projects. I was a little bit strange, maybe. Anyway, I had nothing to do. This one is my first research activity [shows something to Aspray]. This is a planar wave guide, formed by electrically induced migration of ions. These are the graphs, and from here we put in some ions, like potassium or thallium. Then we can make a high-index layer like this. I got this idea from very old literature. This is 1925.

Aspray:

How did you even come across this?

Izawa:

When I finished my doctoral thesis some guy in NEC brought us selfoc fiber. It's some kind of optical fiber. The refractive index is changed gradually from the center.

Aspray:

To the outside.

Izawa:

Yes. It's very special optical fiber, but they invented that. We call it selfoc. He brought selfoc fiber to my professor, and we discussed how to make this. Very strange. My professor knew this paper. This is not related to the selfoc fiber, but I know this process. When I entered NTT and had nothing to do, I thought of the selfoc fiber, how we could connect a signal by the selfoc fiber. After that we needed a planar lightwave guide. This method could be applied to make a planar wave guide. And I succeeded, but this was too early because selfoc fiber is not so good for communication because the transmission loss is very high. But now this method is adopted by Corning. Last year I visited Corning in France, a suburb of Paris, and they are making planar lightwave guide circuits, like a one by eight divider with my method. I told them, this was my sentimental journey!

Aspray:

Had they learned about it from your research?

Izawa:

Yes. He learned everything. He knew my patent, my literature. But unfortunately my patent was already expired.

Aspray:

Yes.

Izawa:

This was my first job in NTT laboratories. With this discussion I learned about the glass, and then I was interested in glass very much.

Aspray:

Did you continue as a researcher in the company to have close relations, close communications with your professor from the university?

Izawa:

At that time, no.

Aspray:

Not at that time.

Izawa:

Maybe I should say yes. Related to this research I had no communication with my professor, but with the gas laser we supported his laboratories, so we had communications or that. One reason I entered NTT is that at that time NTT funded my professor about two million yen per year. For the public university, that money was very big.

Aspray:

Why don't you tell me about your career as it goes along? I'll let you just tell me about the different areas you worked in.

Izawa:

When I was doing this work I knew that the selfoc fiber was not good for communications, so I thought we should make very low-loss optical fibers. I started the research of the optical fiber application. But at that time my knowledge was very poor about glass, so I learned from very thick books. It was not so good for me, and I tried many processes, but I failed.

Aspray:

Were there other people in NTT who knew about this subject?

Izawa:

Nobody. But my boss encouraged me, "It will be very important, so you should do that."

Aspray:

Was there research going on in companies elsewhere in this subject area?

Izawa:

No.

Aspray:

No? It was too early for that?

Izawa:

Too early. But after maybe ten months I knew that some researchers in Corning had done very nice research. They said that they had succeeded in making a very low-loss fiber, of twenty decibels per kilometer. My loss was about one thousand decibels per kilometer, so there was a very big difference. I thought that I'd failed. But the details were not reported, only one or two papers were reported, and especially there was no report on how to make them.

Aspray:

That was considered proprietary?

Izawa:

Yes. So I started again. I used my time very much for the development of optical fibers. I noticed a very old paper from Corning.

Aspray:

James Franklin Hyde, from 1934.

Izawa:

Yes. This said how to make very low-expansion coefficient glass. "This must be very good," I thought, and I tried many methods following this pattern.

Aspray:

This general approach.

Work at Musashino and Berkeley

Izawa:

Yes. The vapor pressure of the dorpant-like germanium oxide was very high compared to the silicon dioxide, so it's not so easy to make this kind of transparent glass, and I failed. At that time many managers in the NTT research labs noticed that optical fiber communication would be very promising, and some guy started research on optical fiber in the NTT Ibaragi Laboratory. At that time I was working at Musashino. It's at Tokyo. Some big boss considered that they should be joined together. But I objected. My house was in Tokyo. My boss told me, "You should go." At that time NTT would send me to some university to do research, and it would be very interesting for me to stay in a foreign country for one year. So I was struggling...

Aspray:

To decide which to do.

Izawa:

Finally I decided to stay here, but my boss said, "Cool your brain (head) in the United States." So I followed that, and I stayed one year at the University of California, Berkeley. I did some research on metal insulators, metal infrared detectors. It's not perfectly related to the optical fiber, but I was thinking how to make it.

Modified Chemical Vapor Deposition (MCVD) and VAD

Izawa: When I returned to NTT, in 1975 I think, I started again, and actually I moved here. There was one more method, called modified CVD method, invented by McChesney, who was a researcher in Bell Labs. This was published in 1974, I think. This is the picture of the patent [shows Aspray], and we have here the silicon glass tube. Inside the glass tube we blow the gas of silicon tetrachloride or germanium tetrachloride, and outside of this tube we heat it up with a hydrogen-oxygen torch. The gas flowing inside changed not only the tetrachloride, but also the oxygen, so this gas will be changed to oxide and deposited on the surface in the wall of the tube. This part will deposit the particle and will be changed to the transparent thin glass layers. So this will rotate and this torch will go back and forth. We call this method the modified chemical vapor deposition method, the MCVD method. This procedure was opened for everybody and is very simple. The fiber made with this procedure is very nice. Japanese people were thinking, "We should have our own procedure, process, or method to make optical fiber." My boss told me, "Okay, we will start the research on this method, and we will make fiber with this method, but you should develop a new method much better than this method." So I started to develop a new method, and finally I developed the VAD method.

Aspray:

Do you want to tell me about that?

Izawa:

Yes, essentially this patent says many things about how to make very nice glass, so I learned from this patent many things, and tried many processes.

Aspray:

This is the 1934 paper you're talking about.

Izawa:

Yes. Finally I noticed that once we made porous glass with germanium oxide, covered the porous glass with pure silicon dioxide glass, and heated it up, the germanium oxide did not go out. It remains in there because it was covered with the silicon oxide. Then we can make pre-form. First I tried to make a very nice shape of porous pre-form. There's a higher index porous glass inside, and the shape should be a very nice circle. The diameter should be very smooth along the actual direction. It was not easy. We needed many controls such as growing speed and deposition rate. The most difficult thing is how to make the torch. There are many secrets. A very simple torch is made with silica-glass, but if you use a regular torch, we put the silicon tetrachloride gas into the torch, so the nozzle of the torch would be blocked with the fine glass particles. It's not so easy to develop a very nice torch for deposition. So I tried many —

Aspray:

Many different variations, yes.

Izawa:

Finally we succeeded in making a torch. Then how to make the transparent glass? It was very difficult. We had just a porous pre-form, very fine glass particles. The diameter is maybe one hundred microns or less. About one thousand angstroms, so one hundred nanometers. Between them there is air, so I developed many kinds of furnaces to consolidate the porous pre-form into transparent glass.

Aspray:

When you were trying these different approaches, did you have in mind techniques that were cost-effective to manufacture in large quantities, or you were just trying to get something to work at this point?

Izawa:

This method is essentially for the mass-production of high-quality fibers. Of course I was thinking about that, but before that it should be high quality. Every day I made a new furnace. My subordinate was working experiments. Maybe in the evening I accepted the report and discussed which ones should be changed. I asked some workers outside to make a furnace for the next morning. I tried everything. I started in winter, maybe in December, and in summer, maybe August, my boss said, "You are doing very nicely, but you have failed. It's not so promising. You should quit that research." I was very confident I would succeed, so I continued, but my mind was not so stable.

Aspray:

Would it be typical in NTT for your superior to tell you that you had to stop work, or did they just suggest these sorts of things?

Izawa:

It depended on the manager.

Aspray:

But you had convinced your manager that it was promising enough that he would let you continue.

Izawa:

No, no, no, no. It's not so simple.

Aspray:

Not so simple.

Izawa:

Two months after when I was told to quit that research, I succeeded. Then everything was okay. But if not, maybe I would have been moved to another strange position or laboratory. I'm not sure! A very small part of the glass was changed to transparent. Then everything went very well. We could make very nice optical fibers compared to the MCVD method.

Outside Vapor Phase Deposition (OVD)

Izawa:

The Corning people developed one more method we call the outside deposition method. This used fine glass particles, deposited on a mandrel. Actually this method was not so good, so they modified it. They mixed it with my method, I think. I'm not sure about that. But my method is essentially like this. We need starting material, just a rod, and at the end of this starting rod we put the fine glass particles from the torch. One is for core glass, and the other is for the cladding layer, and they make a porous pre-form, which is similar to the inter_____. After we made the porous pre-form, it would be heated up to about fourteen hundred degrees centigrade and changed to transparent pre-form. This is very simple and good for mass production. Compared to the Corning method, we call this an OVD, outside vapor phase-deposition method. This method uses also a starting rod, this is just a mandrel, and on the surface of the mandrel they put fine glass particles. After that they take out this mandrel and consolidate it into transparent glass. But this porous pre-form is very fragile; it's not so easy to take out. They don't take out this mandrel. I don't know how they do it.

Aspray:

I see.

Izawa:

In my method, the process is very similar to this Corning process, but my process is simpler and good for mass production. In the case of the Corning process, or the ATT process, there are some index dips in the center because if you take out the mandrel, then diffractive some part of the germanium oxide vapor will go down, so the contents of germanium oxide in the center part will be decreased.

Aspray:

Right.

Izawa:

So there's some dip. It's not so bad, but not so good. In my case, there's no dip. The process is also very simple. That's why many companies use this method.

Aspray:

When you succeeded in developing this process, what were the steps for NTT to put it into use?

Izawa:

As I told you, we don't have any factory, so at that time we had some kind of consortium with wire companies like Sumitomo-Denko or Fujukura Densen and Fujukura Denko. So we showed everything about how to make optical fiber with this method, and they ran everything. They made plants for mass production. Actually we had a very big machine at that time in this laboratory, but we could not sell any fibers, even for our own use.

Aspray:

You can't even produce for internal use?

Izawa:

No.

Aspray:

Is that because of a government regulation about the way you work?

Izawa:

Maybe. I should say so.

Aspray:

How did this optical fiber fit into the company strategy though, for use over time? Was there a goal to put it into use in certain areas, for certain applications, by a certain date? What was the plan?

Izawa:

At that time we could get the MCVD or the ATT method fiber from wire companies, so many people believed that the ATT method would be much better than the domestic method fiber. So there was big competition. At the first stage, most of the fiber used in the field trial of the optical fiber communications was made by the ATT method, and a very small part of the fiber in cables used my method. Just for comparison, you know. But after a while many people believed that my method is good, so they gradually changed the situation.

Aspray:

Are the optical fibers made by the two different methods any different in characteristics?

Izawa:

Almost the same, but my process is very good to decrease the content of water. If there is some amount of water there is a big absorption loss at 1.3 or 1.4 microns, the infra red region, and it's not so good. This method very easily removes the water vapor.

Aspray:

And on a cost basis your method was less expensive?

Izawa:

About the cost, it's not easy to say, because once you have some instrument for, let's say, the ATT method, it's not easy to change the method because you have already invested in one.

Aspray:

A capital barrier to switching technologies.

Izawa:

Yes. It's not easy to compute the cost. In the first stage most of the fiber was from the ATT method, and after that maybe half and half. Right now even ATT people use this method.

Telecommunication systems at NTT

Aspray:

I see. And to what degree are the telecommunications systems of NTT based on optical fiber today?

Izawa:

We have many telephone offices in Japan, and most of them are connected with optical fibers, so I should say the trunk line of telephone line is optical fiber. We use a satellite system, or microwave system, but it's mainly for earthquakes.

Aspray:

To avoid natural disasters. Also, I suppose it's easier to use satellites in certain remote areas where there is rugged terrain, or things like that.

Izawa:

Even the remote islands we already connected with optical fibers. We used submarine cable.

Aspray:

I see.

Izawa:

Right now we are planning to make access lines from telephone offices to homes and offices.

Aspray:

Using optical fiber?

Izawa:

Yes. It costs very big money. My company is now saying it will take until 2010...

Aspray:

To finish this. But I suppose when it is done you can provide much more sophisticated information services to these customers.

Izawa:

Yes.

Aspray:

You say this method has been used in other companies. Is this a good source of income through licensing fees for NTT?

Izawa:

Yes. I don't know how much royalty they are paying to NTT, but they are paying some royalty. A very small amount of money.

Aspray:

I see.

Izawa:

Some small part I can get!

Aspray:

The company must have been pleased with your performance when you came up with this method.

Izawa:

I hope so.

Aspray:

What happened to your career then?

Izawa:

It's not simple. I did three years of developing this method.

Aspray:

When did you finish this?

Izawa:

In 1978. I continued only three years after 1975.

Administrative Work

Izawa: Then my boss told me, "You did a very nice job. After that many people will do that, so you should develop a new fancy photonic device. You should move to the Musashino laboratory." I objected, of course, but at that time, finally I obeyed him and moved to Musashino. So I did only three years more of this research. After that, I started research on the planar light-wave circuit again. It's a different method, applying this porous glass fabrication method, as I showed you before.

Aspray:

That's right.

Izawa:

I continued that research for two years. Then again my boss said, "You should change your job." He told me, "You should do personnel affairs of researchers."

Aspray:

I don't understand.

Izawa:

You know, deciding the salary or bonus of the researchers, or promotions. Everything.

Aspray:

So it's kind of a way of training you to be an administrator?

Izawa:

Maybe. I did two years of that.

Aspray:

I can see that you didn't enjoy that job very much.

Izawa:

I couldn't enjoy it, but I learned very much.

Aspray:

What sorts of things did you learn from doing that?

Izawa:

Essentially I am very shy, and I am not so good at communication. Especially with persons I met for the first time. For me it was a headache to do that. But through that job I learned how to do that.

Aspray:

Did it also give you a better sense of what was going on overall in the research area?

Izawa:

Yes.

Aspray:

Was moving you back and forth between projects and laboratories and cycling you into this job part of the common training pattern in the company to develop its researchers for more senior positions?

Izawa:

Yes.

Aspray:

Is this something that you are instructed to do as a manager today?

Izawa:

Yes. But I'm not sure this procedure is good.

Aspray:

I see.

Izawa:

Of course, it depends on the person. I know many researchers passed a similar career. Maybe I shouldn't say so but most of them failed in the process.

Aspray:

But it may be a good filter.

Izawa:

Yes, good for filtering managers from researchers.

Aspray:

I think it is very hard to identify who would make a good manager out of a research organization.

Izawa:

Yes.

Aspray:

What did you do after your two years in this personnel position?

Izawa:

After that I got the section head of the silicon and gallium arsenide crystal growth laboratory. I don't know anything about that, I was just the manager, and I continued for two years of that. After that my boss told me, "You should move again." A new section was super-lattice. Do you know super-lattice? It's very thin crystal layers.

Aspray:

I see.

Izawa:

With that crystal we can make very high-speed electronic devices, transistors, ICs, or very nice lasers. I was head of that section in the laboratories.

Aspray:

As the head of one of these sections, what are your responsibilities?

Izawa:

I had about twenty researchers at that time.

Aspray:

Were you expected to be in the laboratory with them doing work?

Izawa:

No. Just managing.

Aspray:

Just managing. So you had to see that schedules were maintained, that the right personnel were there.

Izawa:

Yes, getting new researchers, or extra research money, or telling researchers who are doing unpromising research to quit.

Aspray:

I see. Did you have to make that evaluation yourself?

Izawa:

Yes. The evaluation of research is not easy, but someone should do that, otherwise the research area will grow and finally go bankrupt.

Research Director and Move to Ibaragi

Aspray:

Yes. Then you moved?

Izawa:

Then I moved to director of basic research at the laboratories. There, the researchers are doing many things. Some guy is managing Schroedinger equations, to explain why we can get high-temperature superconducting. Or some guy is taking out a cell from a mouse's brain and doing something. Or some guy is synthesizing some new polymers of silicon. Regular polymer is made of a carbon chain. Instead of a carbon chain they are developing silicon chain polymers. It's very interesting, but just research right now.

Aspray:

This basic research, was it all on the materials side, or was some of it systems work also?

Izawa:

My lab was just the materials side.

Aspray:

Okay.

Izawa:

At that time about one hundred researchers were working under me. I worked for six years as that. It was my longest time in one position! After that I came back again to Ibaragi (this place) to do my own research. The research field is my own...

Aspray:

Is your own special area. Do you think that being in a research organization where you are fairly knowledgeable about it makes you a better manager?

Izawa:

I think yes. But we should be careful when we say something to researchers, because I know everything about old research, and my idea will be affected by old research. So if a young researcher says some new idea, it's very easy to say, "No, no, no, it's impossible. It's difficult." So my policy is that I never say, "No. You shouldn't do that." Instead of that I say to the young researcher, "Okay, you should try it for, let's say, half a year. One year." Or, "How long do you need? After doing some research, then decide." I think it's one of the best managing methods for developing new materials or new devices. The manager should know everything about the research field, but the manager shouldn't say, "No," at the first stage. It's my policy.

Aspray:

I see. In terms of choosing the areas for research, how are the decisions made? Do you regard some of the other laboratories within NTT as your basic customers, who define your research agenda, or do your researchers come up with the ideas and then you tell the other laboratories what they think are promising results? How do you define your research agenda?

Izawa:

We have many procedures. Of course, sometimes I ordered, "We should do this kind of research, we should do that." In that case I make a small group to do that. In some cases a researcher proposed to do the research. We do it both ways. Bottom up and top down.

Aspray:

When you impose upon your researchers a particular project, how do you come to the idea? Does somebody impose it upon you, or is it a good idea of yours? Just how does it happen?

Izawa:

When I ask to do some research in my laboratory, I don't know the solution, but I know that we need this kind of device or material, so we should develop it. Of course, personally I have some kind of idea about that, but I don't tell the researcher before, you know. After a while, sometimes I tell him my idea.

Aspray:

Do you have regular meetings with other laboratories? Will the managing director of another laboratory say, "Here is one of my problem areas," or, "I'm trying to build a system that does this, but I don't have the materials or the device to do that." Is that kind of information passed?

Izawa:

Yes. We have many kinds of meetings! More than enough! As I told you, we have twelve laboratories, so every week we have a meeting with the executive manager of every laboratory. And the location is appointed.

Aspray:

You go to one place one time and another place another —

Izawa:

No, no, no. We have a TV meeting system.

Aspray:

I see. So you don't have to travel every week.

Izawa:

Sometimes we meet about some special topic, like the optical fiber access network. We have some special topic meetings, which will be held maybe once a month, or something like that. Maybe three times per year for some topics. Researchers communicate with each other maybe every week. It depends.

Aspray:

There are technical committees for the researchers also?

Izawa:

Yes.

Aspray:

In addition to meetings, how do you spend your typical working day? What kinds of things do you work on?

Izawa:

It depends on the season. Today is May 17th. The fiscal year was ended March 31st, so at the end of next month we have a meeting of stock owners.

Aspray:

Stockholders.

Izawa:

We have some change of big bosses. Not everything, but some part of the big bosses might be changed, so according to that change we must rearrange organization personnel. Right now I'm thinking what should I do with certain guys. Maybe I should fire them or move them to another section.

Aspray:

So you're thinking about possible things you might do if you have to reorganize, or if your budget is cut, or your budget is raised.

Izawa:

Maybe in January, or maybe in December I was thinking about the research budget. All research should use more money to accelerate the speed of development.

Aspray:

For the coming year?

Izawa:

Yes, before the next year.

Aspray:

That starts in April?

Izawa:

Yes. And in the autumn, maybe September or October, I'm thinking of the rating of the project or of individual researchers.

Aspray:

Do you do a personal evaluation?

Izawa:

Yes.

Aspray:

Of every researcher?

Izawa:

No.

Aspray:

Or just the managers?

Izawa:

Just the upper-grade managers.

Aspray:

Then they review each one below?

Izawa:

Yes. Actually I do two layers under me.

Aspray:

I see.

Izawa:

In addition to that internal business I must do at some point outside business, such as international conferences. Committee meetings of academic associations such as IEEE, or in Japan we have also some organizations such as IEEE.

Aspray:

It's expected that your researchers will be active members of that community?

Izawa:

Yes, that's right.

Aspray:

Including the managers?

Izawa:

Yes.

Peculiarities of Research Administration in Japan

Aspray:

How much of your time is taken up with ordinary business issues of running a research operation? Regular financial things, personnel matters, and so on?

Izawa:

Eighty percent I think. Of course, it depends on the time. Right now I am co-chair of an international conference that will be held in next year in Hong Kong.

Aspray:

And that has taken a lot of time?

Izawa:

Yes.

Aspray:

How do you get your budget?

Izawa:

About the same budget as last year, I can get without a big battle. But in these last two or three years the economy in Japan have been...

Aspray:

Very difficult.

Izawa:

The income for my company maybe is not so good, so there is a small struggle with the R&D. They decide how much money we can use. So for two or three years my research budget has not increased, but we have many requests to develop new devices and new materials. People say, "We don't have money anymore." I say, "I need more money!" I think we have a battle between our laboratories.

Aspray:

I know one research organization in the United States where they bring all the different managers of the different research units together in one place, and they all make presentations one after another and they fight with each other to decide who gets how much money.

Izawa:

No, we don't have such severe battles!

Aspray:

That's very good.

Izawa:

Many people want to get very nice photonic devices in the future, so nobody objects to using money for that. So the situation is not so bad. The economics of these last two or three years was not so good, so that's one major reason why we struggle about the research budget.

Aspray:

Do you have to make a formal presentation to the senior researchers when you make your request for your budget?

Izawa:

Yes, but it's not effective. Just like a ceremony. Actually the struggle is between me and the staff of the R&D headquarters. It's a big pattern of Japanese struggle. It's not an open place, but between two persons. This is the staff of R&D headquarters, I ask him, and he thinks how to get money for me, and he discusses it with other research managers.

Aspray:

What about personnel issues? For example, how do you recruit new people? How do you recruit people from other parts of the company? How do you make sure that you don't lose your good people to other parts of the company? How do you keep your researchers happy in their jobs? How do you shift them to other things when they don't want to be moved? Can you talk about some of those issues?

Izawa:

Most of the researchers working in NTT are fresh from the university. In the research lab they come mainly from a master's course. Most of them will work until the age of fifty. Some of the researchers coming to NTT laboratories from a university: An assistant professor position, post-doctoral position, or something like that. Just recently we started to hire from foreign countries, including the United States, and Germany. Most of them are in post-doctoral positions for one or two years, but some of them already have a permanent position.

Aspray:

I see. Do these people have to speak Japanese?

Izawa:

No. They learn Japanese, but in daily life in the laboratories they speak English.

Aspray:

I see.

Izawa:

We Japanese speak very poor English, so the foreign researchers will improve our researchers' English. I hope. Especially in the basic research laboratory we have many foreign researchers: Indian, Chinese, Korean, American, Europeans; more than twenty. It's quite different from the United States in that most of the researchers do not change companies in the young range especially. Some of them change jobs from here to, for example, a university. Most of the researchers want to get a professor's position, so, for example, my laboratory has two hundred and fifty researchers, and last fiscal year about six researchers quit.

Aspray:

To take professor's positions?

Izawa:

Yes. And there are some special cases. We have a very nice researcher, of the basic research lab, and Stanford University wanted him to be a tenured professor. He wants to be a professor, but NTT objected, so the final solution is he can get a salary from Stanford and also from NTT. He works nine months for Stanford and three months for NTT. So he has two positions.

Aspray:

Yes. How do you keep researchers from getting stale in a particular position?

Izawa:

Especially right now, compared to the other laboratories, including in foreign countries, like IBM labs, or Bell Labs, the facility or research budget of my laboratory is very good. This is not only my opinion; researchers who finished here told me, "Your lab is very nice. The facilities are good. The salary is not so bad. The budget is good." So if one guy is a good researcher, I think, he does not want to quit NTT. I am very confident about that.

Aspray:

But sometimes it is maybe good for the company to switch employees from doing one kind of project to doing another kind of project, the way you were encouraged to switch tasks. My impression is that researchers don't like that very much.

Izawa:

Yes. In some cases it will happen. One example is about ten years ago with the high-temperature superconducting material. It's some kind of oxide. We got a fever of that research, so many researchers moved to that field. But right now the fever is cooling down, so except to the very good researchers, a manager will say, "You should leave that field." But some guys insist on continuing that research. In some cases researchers quit NTT and moved to universities or some consortium research lab. I don't remember the names, but they are very rare cases.

Aspray:

Does your entire budget come from within the company, or do you get some grant money from other government agencies?

Izawa:

We get some budget from the government, but it's a very small part. Rare cases. Now we have maybe one project from the government. The money is very little compared to our internal budget. So we don't care about the government budget. It is just co-operation.

Aspray:

You're doing this to be a good part of the research community for your country?

Izawa:

That's right.

Aspray:

To what degree are there consortia of researchers or representations from different companies and different universities in a particular area? Say in the materials area. Do they have some impact on what your program of research is? Suppose, for example, there is research going on in photoelectronic devices in two-dozen organizations in Japan. Is there an over-riding professional committee that brings all those groups, representatives of all those institutions together?

Izawa:

We have many societies like IEEE, and in that society we have many groups, like LEOS. There are many, many groups. Using that society we have many meetings to exchange ideas or report on research results.

Aspray:

What about government-organized committees? For example, yesterday when I was doing an interview on electron microscopes, there was a committee in Japan of the various groups doing electron microscopy, and they set the research standards for all the Japanese researchers in this area.

Izawa:

Standards?

Aspray:

Not a standard in the official sense, but they would agree among themselves that here are the central problems to be worked on, and we don't have all the resources to do everything ourselves, so this organization agrees to do this part, and this other organization agrees to do this other part. They break up the project into pieces.

Izawa:

I see.

Aspray:

Does that happen in your field very much?

Izawa:

No. We don't have such a pattern. Of course, we need a standard, like for optical fiber connectors. Not just any kind of connector can be applicable. In that case I developed and decided on what diameter, what shape, what's important. So we decide the specification, we try to improve what's wrong and modify it, and finally we ask to outside companies to make them. With that we try to get that connector to be a worldwide standard at the meeting of the CCITT, or ISO, or some other. There are many kinds of standards organizations; we try that. We spent money for that. Standards are very important for communications.

Aspray:

What about joint research? Do you have research with other companies, or with universities?

Izawa:

Yes. Actually it's not joint research, but we support universities not only in Japan but also in the United States and some European countries.

Aspray:

Do you have certain rights to research results in those cases?

Izawa:

Yes. Formal letters state that the patents should belong to NTT. It depends on the situation and how much money we give in support. But in a typical case, that kind of joint research is in the area of very basic research, a, patent is not..

Aspray:

Is not so common. Yes.

Izawa:

Even if they get some patent, it does not make sense.

Aspray:

Yes, I see.

Izawa:

We send many researchers outside, for example to American universities, to stay one or two years for research. We have two purposes. One is to get some experience staying in a foreign country and speaking a foreign language, such as English, German, or French, and to do some research at that place. Their way of research is a little bit different, maybe, and to get good friends in the research field.

Aspray:

Yes.

Izawa:

In some cases we get very nice results. For example, just recently we got a very nice result with a 1.3 micron amplifier. The 1.5 micron wavelength amplifier is now popular, but the 1.3 micron wavelength amplifier is not so easy. Some guy noticed that some material would be very good for 1.3 microns when he was working at the Rutgers University.

Aspray:

Oh yes, Rutgers university.

Izawa:

Professor Shnitzer, he is a very famous professor about glass, and some researcher spent one year there at his laboratories and did very nice work.

Aspray:

I see.

Izawa:

Maybe discussion with Professor Shnitzer was very effective for that person. Mainly in such cases we have many joint researches. Every two or three researchers will be sent outside. Maybe one percent of those researchers will do very nicely.

Aspray:

That's a lot of people.

Izawa:

Yes.

Relationship between NEC and NTT

Aspray:

I know that in the past, and it may still be true for other parts of NTT's business, that NTT and NEC have had very close working relations on certain research areas. I don't know how those operate. Can you tell me about either that one or some others?

Izawa:

I don't know the details about NEC and NTT relations, but I can guess. When we want to develop some new equipment, such as a telephone exchange machine, our researcher designs the outline of the exchange machine. This part should be like this, and this should be like this. Something like specifications. He asks if NEC people can do this or not, and modify it. After some discussion NEC will make a real exchange machine and we accept that machine, check everything, and modify it again. That's a typical way of past development procedure.

Aspray:

I see. Does this particular research division have certain kinds of relationships with outside companies that specialize in materials, to develop something jointly, or have a working relationship, or contracts with them to do research?

Izawa:

Yes. In the past year, in the case of the Vapor-phase Axial Deposition, the VAD, it's the same as the relation between NEC and NTT. We had a consortium with three major cable manufacturers and NTT to develop the very nice optical fiber cables. Now we are thinking about nice material for optical wave guides, made by some kind of polymers, because with polymers we can get a very cheap, low-cost wave guide device. We synthesized a new material, but it's not easy to get a large amount of such a material. Under a contract we showed everything and asked to make that for some amount of material. But it's a very basic research case, and right now we have a typical procedure to get a joint research company. We call it Track Three. That system was made mainly because of pressure from the United States. As you know, Japanese companies are not so open to foreign countries, so NTT decided to take on some open route to make joint research consortia. We opened up what we wanted. Not detail, but a rough schedule. A company outside makes some proposal related to that opening, and we decide on one or two companies. For example, we are now developing the ONU. It's a small box for the Optical Network. ONU means Optical Network Unit. It is for the office or the home, to terminate optical fibers. To develop that we used Track 3, ATT, Hitachi, and Fujitsu passed that system, and it is almost finished. So right now we have a very fair, open system for development of a new device or system.

Linkages between Research and Management Positions at NTT

Aspray:

In the culture of NTT, is it possible for people from the research organizations to move into very senior management positions in the company? General management positions.

Izawa:

Yes. My boss was a researcher, and now he's a senior vice-president. He's a member of the board of NTT. We have about thirty persons on the board.

Aspray:

This is an internal board of staff?

Izawa:

Internal board. In Japan it's very rare to have an external board member. Most of them are internal board members.

Aspray:

Yes, that's much more like the European system than the American. But is a board member's responsibility still for research?

Izawa:

Oh, yes.

Aspray:

Is it likely that such a person would move into a staff position, or into central general administration?

Izawa:

Yes. It's not popular, but for example, Doctor Bun'ichi Oguchi is very senior person, and he was second from the top of NTT. He was a researcher, of microwaves. But it's not so popular.

Aspray:

In the United States even the most technological businesses have lawyers and financial people at the head of the corporations, but I'm not so sure that that's the case in Japan.

Izawa:

Yes. Anyway, it's my personal opinion. For example, in an electric power utility such as Tokyo Denryuku (TEPCO), or a steel company — a basic industry, lawyers or business school people will be at top.

Aspray:

Appropriate to be at the top.

Izawa:

Yes. But I think the president of the NEC is an engineer. Many electrical or electronics companies, like Hitachi, Fujitsu, or Toshiba, have many engineers that are not researchers. Engineers will be on top in that company.

Aspray:

Which is different from the United States, because IBM, General Electric and Westinghouse are all run by people from the business school side, not from engineering backgrounds.

Izawa:

I am not sure which one is better.