Policy Study Compares Nanotechnology Funding in United States and Japan
Burgess Laird, a Senior Policy Analyst at the Center for National Security Studies, Los Alamos National Laboratory, is completing a study comparing the funding situations for nanotechnology in the U.S. and Japan. He presented an early version of the study at the American Society of Mechanical Engineers meeting in December 1993, entitled "The Political Economy of Nanotechnology: A Comparative Study of Japanese and U.S. Policies, Supporting Institutions, and Competitive Prospects in an Emerging Strategic Technology." A revised version is now being prepared for publication. The views presented are his own and do not represent those of Los Alamos Laboratory.
The author was interviewed at the ASME meeting by Foresight editor Chris Peterson.
Foresight: You've been studying nanotechnology research funding in the U.S. and Japan. How would you compare the levels of commitment?
Laird: If the research projects that Japan and the U.S. have now in place in nanotechnology and nanotechnology-related areas are any indication, then I would say that Japan is making a much bolder investment in nanotechnology R&D. The boldness of this investment is reflected in an expenditure of roughly $560 million over the next ten years.
It's also reflected in the fact that it's a ten-year commitment. Two MITI programs alone account for $240 million worth of work. And it's also indicated by the fact that Japan has gone so far as to reorganize their national laboratory system in Tsukuba Science Cityat least those laboratories that are MITI-related under AISTaround the "Ultimate Manipulation of Atoms and Molecules" (UMAM) project, which is Japan's lead project in nanotechnology.
The lead project in the U.S. is ARPA's project called ULTRA, standing for ultra-fast, ultra-dense computing components. It's primarily electronics oriented, and their approach is far more "top-down" than "bottom-up," though there's a significant thrust in the area of chemical self-assembled structures. Japan's new UMAM project, as well as ongoing ERATO programs, appear to stress bottom-up approaches as much as they do top-down approaches.
Foresight: What are the two MITI programs that you mentioned?
Laird: There are two large MITI efforts: one is related to nanotechnology and the other is micromachining. The Japanese schedule one project after another. One started in 1991; that's the micromachining work. The next is the UMAM manipulation project started in 1992. They are both ten-year, 25 billion yen efforts, worth approximately $227 million per project at 110 yen to the dollar.
The micromachining effort, which is mostly MEMS [microelectric-mechanical systems] work, also puts an emphasis on bottom-up approaches which should be useful in developing molecular nanotechnology.
But there are several other nanotechnology projects aside from the MITI projects. There are several ERATO projects, three of which are ongoing right now. Each of these are five-year, $10 million projects. The best known is the Aono Atomcraft project.
There is also some work being sponsored by RIKEN (Institute for Physical and Chemical Research) in three separate laboratories, which in terms of scale are probably comparable to the NIST projects in the U.S. And the Ministry of Education has been funding nanotechnology projects for several years as well. So those are the main projects in Japan that are going on right now.
Foresight: It's interesting that their micromachine effort has some bottom-up components to it. I hadn't realized that.
Laird: It is; but they are using chemical synthesis techniques to create microscopic flagellar motors.
Foresight: So in Japan we see one big bottom-up nanotechnology project, and another one which you would think would be strictly top-down MEMS work, but where they are doing some bottom-up work as well. While over here in the U.S., there is the ULTRA project, which you would hope be doing something with the bottom-up strategy, and they're focusing on top-down.
Laird: By and large, that's true. Within the ULTRA project there are also some efforts in cellular automata, which I know includes a lot of work in protein folding and structural biology, or at least the simulation thereof.
Foresight: But what about chemistry and proximal probes?
Laird: As I mentioned, ULTRA does have a thrust in chemical self-assembled structures, but most of the work seems to be in top-down materials and fabrication technology.
Foresight: Is there anything that looks like the development of molecular machines?
Laird: I am not aware of any significant efforts in the United States.
But let me tell you some very interesting things about the Ultimate Manipulation project.
First, in contrast to the largest U.S. effort, the UMAM effort is explicitly oriented to achieving civilian as opposed to defense applications. Second, the majority of the UMAM efforts are going to be carried out at one placethe National Institute for Advanced Interdisciplinary Research (NAIR) in Tsukuba Science City. That is a novel departure for Japan. They've always had government/university cooperative efforts. They wrote the book on that.
They use the term "condensed research" to describe what they are doing at the NAIR facilities in Tsukuba. I would describe it as interdisciplinary basic research designed for speed. What else is the motivation for this centralization? I certainly don't know what other motivation there is for coming together in one physical location, in this age of distributed computing and networks, unless you want to move fast and get as much synergy as possible, using all the personal networking possible in one place.
Foresight: Would you say that it was the Ultimate Manipulation goal that stimulated the reorganization of their research institutes?
Laird: Well, the reorganization preceded NAIR's establishment by only one year. The reorganization was announced in January 1992, and then in January 1993 NAIR was set up. I don't believe it was until June 1993 that all the pieces were finally in place.
I should also point out that MITI's micromachines and UMAM projects constitute the latest "national, large-scale" research projects in Japan. So aside from the fact that it's "condensed style," the Japanese see this Ultimate Manipulation project as their "leading edge" thrust, if you will, into large-scale, nationally-coordinated basic research.
Of course, ERATO has been doing basic research for years, and the Ministry of Education has been funding university researchers in basic areas.
At this point 79% of all the funding in nanotechnology comes out of MITI. There are other MITI programs that are related. For example, quantum functional devices. This year's funding is 700 million yen, roughly $6 million; not peanuts. Still the QFD effort is not a part of UMAM.
Foresight: About your study in general: did you initiate this project or was this requested by Los Alamos?
Laird: I did this, not as a result of direct project funding, but as an analyst in one of the Lab's indirect-supported Centers. My goal was to conduct a case study comparing U.S. and Japanese support and organization of long-term research in a cutting-edge field.
Foresight: Do you think thatnow that you've done this workthere will be people at Los Alamos and elsewhere who will be interested in seeing it?
Laird: I always hope that my analyses will be useful to long-range planners. Nanotechnology, after all, is a long-term, high-risk undertaking that firms on their own will not readily undertake because of the cost, and because if they should realize any breakthroughs in it, those resulting discoveries would be quickly diffused. It's a classic public good, market failure.
It would also appear to be a strategic technology that, in time, could impact many fields and industrial sectors. Its advancement is going to be propelled by progress on many fronts, from computation to materials, and that is going to require integrated networking of these advances.
Foresight: How did you originally get interested in this subject?
Moreover, I've been interested in strategic technologies for some time now, because they are areas at the intersection of science, technologies, policy, and international affairs. I look at national needsnot technology firstbut national and international needs. My approach borrows insights from several disciplines. I describe my work as political economy as opposed to econometric model building, which by and large treats technologies more from the sidelines, if it treats them at all.
I have an interest in looking at what mechanisms and institutions nations employ to suport science and the development of technology, which are intricately linked to economic growth. So I developed an interest in strategic technology, because it's an area whose pursuit profoundly affects the economic future of nations.
Prof. Requicha is the Director of USC's Programmable Automation Laboratory [more recently, the Laboratory for Molecular Robotics]. He is best known for his work on geometric modeling of 3-D solid objects. The theory and algorithms he has developed are used in various geometric modelers, and several commercially available CAD/CAM systems have been built upon his work.
Dr. Requicha's primary research interests are in automation and robotics, and encompass the design, implementation, and theoretical underpinnings of CAD/CAM systems for mechanical products. With his students he is building novel software systems that exploit and integrate techniques from the fields of artificial intelligence and computational geometry.
More recently, he has taken an interest in nanotechnology and molecular manufacturing. Foresight Update editor Chris Peterson interviewed Dr. Requicha when he was in the Bay Area recently meeting with Foresight chairman Dr. Eric Drexler and IMM advisor Dr. Ralph Merkle of Xerox PARC. Herewith some excerpts of the group's conversation:
Foresight: How did you become interested in nanotechnology?
Requicha: It's a long story. I first got interested in small things in general, and that was mostly due to my friend Peter Will, who used to be up here at HP and now works for ISI at USC. We started talking about micro things and so forth, and from micro we went smaller and smaller. We started thinking about doing things like reading DNA and said "This looks like fun." So what do you do when you're a professor and you want to learn something? You teach a course on it. That works pretty well.
Before I had the courage to do that I ran into Eric's book Nanosystems and I thought, okay, in case of trouble I have a book. Without this I wouldn't have the guts, because otherwise what do you do to prepare a course in a new field? You sift through piles and piles of literature trying to find out what's relevant. It's a very difficult thing to do. So the book gave me enough courage to teach the course.
We also are trying to start research in the area. We've put together a team starting with myself and Dr. Will, who are both mostly robotics and computer science people. We also have surface chemists and material scientists, and we'd also like to have someone from biochemistry. One possible source of funds is a proposal we recently submitted for internal funding from foundation money at USC. There is a lot of competition for these funds from other groups who are all trying to do unusual new things as well. So far we've made it through the first cut. There were about 25 proposals and we made the final five, and we are now waiting to see what happens. So that's how it all started.
Foresight: The course has just begun, is that correct?
Requicha: The course began mid-January and unfortunately, due to some
unexpected events, I was able to do very little preparation. But so
far it's going fine, and I'm staying a couple of lectures ahead of the
Foresight: What kinds of students do you have in the class? Undergraduates?
Requicha: No, graduate students. This is being offered as a special topics course in computer science, which is the easiest way for me to offer a new course. I can just label it "CS 599" and teach anything I want, within reason. I called this CS599 "Molecular Robotics".
Drexler: Nanosystems grew out of a similar course at Stanford in spring 1988. I wanted to learn a subject, so I taught that course and began writing the book.
Foresight: So the USC course is a computer science course?
Requicha: It's offered through computer science, so most of the students are computer scientists. I have a number of electrical engineers, one or two industrial engineers, and a few materials science people in it as well. What I've found is that the computer science students have a difficult time with the physics and chemistry. I told them to go read Part II of Nanosystems and they said, "Oh boy we don't understand anything," so what I'm doing now is teaching principles of quantum mechanics and chemistry.
Foresight: So you have to teach some background?
Requicha: I think I have to teach some background. Otherwise, when you talk about sigma bonds and so on, the students look at you and say, "I want out of here." It's not surprising.
Foresight: What is your own background? You must have a broad background, not computer science alone.
Requicha: In my previous incarnation I was an electrical engineer, and before that I was sort of a physicist, and I did teach physics briefly in Portugal. Then I became a researcher in communications, got tired of that, and became a computer scientist. For the past 20 years I've been doing a lot of geometric modeling, three-dimensional modeling and graphics mostly, for CAD/CAM and robotics. The last few years I've been mixing computational geometry and artificial intelligence to do spatial reasoning.
So my take on all this is that both Peter Will and I convinced ourselves that robotics had something to do with nanotechnology. For example, if you do things based on STMs and AFMs it's easy to see that they have a lot to do with robotics. An AFM is just a manipulator -- sort of an odd little one, but it is a manipulator -- so in our research proposal and also a bit in what I've been teaching, we've been talking a lot about STMs and AFMs and how to actually manipulate things.
The materials science person in our team has STMs in the lab and experience in these things, so for us that's probably the easiest way to get into the nanotechnology field. We like theory, but we also like to make things actually move and happen in the real world.