Appendix C. Site Reports

Site: AIST/MITI Headquarters
1-3-1, Kasumigaseki
Tokyo 100, Japan

Date Visited: September 27, 1993

Report Author: J. Giachino



J. Giachino
H. Guckel
B. Hocker
G. Holdridge
S. Jacobsen
R. Muller
L. Salmon
C. Uyehara
K. Wise


Larry Weber National Science Foundation (Japan)


Hiroshi Kasai Director, Machinery & Aerospace R&D
Satoshi Ito Senior Researcher, Machining Technology Div.
Manufacturing System Dept.
Mechanical Engineering Laboratory


The role of the Ministry of International Trade and Industry in promoting the industrial and technological competitiveness of Japan goes back at least to the early part of this century, but its role has been changing dramatically in this decade. MITI policy now focuses on international contribution and cooperation. The Agency for Industrial Science and Technology was formed within MITI in 1948 to manage a number of government laboratories and institutes (some of which have a history dating as far back as the late 19th Century). Since 1964, AIST has been involved in the promotion of R&D projects through government-private sector cooperation. This has been carried out through a series of national projects, the earliest of which (1966) was dubbed the "Large-Scale Project (National R&D Program)." Several such projects (e.g., Moonlight Project, Basic Technologies for Future Industries, etc.) are underway at any given time; under each project are specific R&D programs.

In 1991, MITI established a program to develop micromachine technology under the auspices of the Large-Scale Project. Since then, the Large Scale Project has been restructured into the "ISTF" (Industrial Science & Technology MITI's Micromachine Technology Project that is of interest to this panel is one of several programs funded under this new ISTF Project.


Structure of Program

MITI's micromachine technology R&D program was initially budgeted in 1991 as a ten-year program with total projected funding of 25 billion. The first five-year phase of this was projected at 10 billion, to be followed by an interim evaluation period and a second five-year phase budgeted at 15 billion.

Our host, Mr. Kasai, stressed that these funding figures are targets. In fact, the 1993 AIST brochure (handed to the panel) shows a total of 2.43 billion actually allocated to the micromachine program in its first three years of funding. Mr. Kasai also said that only 10-25% of the total micromachine budget is devoted to research on approaches to micromachine construction that employ semiconductor lithography and related technologies, which are the focus of this study.

The structure of this program is shown in Figure 7.8. As indicated in the figure, three national laboratories are funded directly by AIST under this program: the Electrotechnical Laboratory, the Mechanical Engineering Laboratory, and the National Research Laboratory of Metrology. Industrial participation in the program is managed through the New Energy and Industrial Technology Development Organization (NEDO) under AIST. NEDO in turn contracts with the Micromachine Center for the management of the individual research projects. As shown in Figure 7.9, MMC funds 23 Japanese companies and one Japanese Institute to perform R&D under this program, as well as one U.S. company, one U.S. non-profit and one Australian university. All 27 of these are considered "research supporting" members. Each research supporting member pays a membership fee which is used to pay MMC's overhead. The JTEC panel learned later that MMC also is considering other uses for these membership fees (see site report on MMC by Linton Salmon -- p. 158).


Under the two-phase approach MITI has adopted for this program, the first five years will be devoted to the development of enabling technologies through a program focus on one or more "model" applications, which may or may not turn out to be realistic. The results of these model application studies and the technologies developed in support of them will be evaluated after five years. The current model will be reviewed and R&D targets will be established for the second five-year phase.

The focus of this program is on applications development, without any commercial level consideration of fabrication approaches. Thus it is not surprising that only a relatively small percentage of the effort is being devoted to semiconductor lithography-based approaches.

The JTEC team's AIST hosts made it clear that they are looking for new ideas for practical applications of micro-machines, as well as new mechanisms for seeking out new applications ideas. The ISTF handout MITI provided to the panel listed "determination of appropriate applications" as a key future subject for the micromachine program. One idea mentioned by Mr. Kasai is the exchange of information and ideas among individual researchers, regardless of seniority, for their ideas on applications. This would be in keeping with MITI policy making, since MITI programs traditionally have been advised by informal committees composed of a variety of outsiders.

Power Plant Application

In March 1993, AIST selected its first model application for development under the MITI program: an advanced maintenance system for heat exchanger tubes in electric power plants (Figure 6.5). This system is envisioned as consisting of several miniature robots: a "mother ship" capable of traversing the inside of pipes of 10 mm diameter (Figure 7.2; a "microcapsule" of 2 mm diameter (Figure 7.3) that would float on the coolant stream in smaller pipes, detecting cracks or corrosion on inside pipe walls; a wireless "inspection module" (Figure 7.4) of 2.5 mm diameter that would move through heat exchanger pipes with an inching worm drive mechanism; and an "operation" or "work" module (Figure 7.5) that would repair flaws detected by the microcapsule and the inspection module.

The mother ship carries the inspection modules and operation modules close to the portion where something is wrong. Then the mother ship supplies power to both modules and transmits the information. To develop this micromachine, R&D will be carried out on: electrostatic driving mechanisms, optical scanners, pneumatic clampers, microbatteries, coupling mechanisms, and group control and behavior control systems. The microcapsule, floating on the stream in the pipes, detects cracks or scales on the inside of the pipe's wall and reports the location to the control center. This will require R&D on: microdynamos, steering mechanisms, microgyroscopes, flaw detectors, and signal transmitters. The wireless inspection module travels from the mother ship to the portion where something is wrong. Then it precisely inspects and analyzes the condition of the changes and reports the results to the control center using the mother ship. For this micromachine, R&D will be needed on: inching worm driving mechanisms, piezoelectric motion drive devices, CCD microvisual sensors, ultrasonic microsonar devices, microphotospectroscopy devices, and teleoperation and coordination control systems. The operation module travels from the mother ship to the portion where something is wrong. Then it repairs the problem according to the report from the inspection module. R&D will is underway for this machine: gear train locomotion mechanisms, multi-degree of freedom manipulators, environment recognition devices, end effectors, photoelectric generators with boosters, high power devices, and valves.

Since the focus of the MITI program in this first model applications phase is on the development of the enabling technologies, perhaps the most significant elements of Figures 6.5 and 7.2 - 7.5 are the key components and technologies that are listed on each figure. As the JTEC panel learned in more detail at MMC, each of these components has been assigned to one or more of the research supporting members of MMC.

It was evident from this presentation that a lot of thought has gone into the powerplant application. Yet our AIST hosts stressed that final specifications of these devices (listed in the figures referred to above) have yet to be determined. Furthermore, by designating this only as a "model" application, AIST recognizes that it is quite possible, if not likely, that this system will never actually be built.

Research related to this model application is supported in part by the Japanese government's electricity use tax.

Medical Application

AIST is in the process of developing a similar concept plan for a biomedical model application, tentatively dubbed "Intraluminal Diagnostic & Therapeutic System." This concept envisions the development of microcatheters in the 1.5 mm diameter range that could be used to treat, for example, cerebral aneurisms and/or blockages of the pancreatic bile duct. Figure 6.6 shows the general concept of this application. Figures 7.6 and 7.7 are conceptual drawings of the outer duct and the inner duct of such a catheter, respectively. According to some of the literature handed to the visiting JTEC team, this application is not as far along in development as is the power plant application. Mr. Kasai stated that this application is now in the final stages of concept development.

Possible Future Application -- Targeting Energy Savings

AIST is now considering the possible addition of a new model application targeted at energy savings in manufacturing. As one example, Mr. Kasai cited the large amounts of energy that are required to construct and maintain big clean rooms for semiconductor manufacturing. He expressed the hope that energy could be saved through the development of microrobots that could aid the manufacturing and assembly of semiconductor devices in clean room facilities substantially smaller than those commonly used. It became evident during the JTEC team's visit that our hosts are just now beginning to flesh out this idea.


AIST is considering a new initiative aimed at addressing the question of how to evaluate the performance of micromachines. Prof. Miura of the University of Tokyo will head up a committee to consider these "evaluation technologies" (metrology issues).


As listed in Figure 7.9, one U.S. company, one U.S. non-profit organization and one Australian university are involved in the micromachine program. Later at MMC, the JTEC team learned that Karlsruh in Germany is also participating in this program, though not as a research supporting member.

MITI initially announced this program as an international effort, and went to some lengths to encourage participation by overseas researchers. However, a number of U.S. firms that did apply to participate as research supporting members (i.e., to be funded to do research) evidently were not able to work out contractual arrangements with MITI that were mutually satisfactory. Under MITI contracting rules, for example, 50% of all royalties resulting from research funded by MITI belong to the Japanese government. Furthermore, all funded participants are required to follow the Japanese governmental budget control systems. During discussion, Mr. Kasai mentioned that AIST is currently reviewing its contracting procedures in an effort to find ways to encourage greater international participation. Professor Wise, the JTEC Panel Chair, mentioned during this discussion that there is a need to share information internationally on the properties of new materials used in MEMS devices. Mr. Kasai responded that AIST would consider requests for information sharing on a flexible case-by-case basis. Mr. Kasai also mentioned that he will be representing Japan at an international conference in October 1993 to find ways of carrying out the International Advanced Robotics Program that was agreed to, in principle, at the Versailles Summit.


Following are answers to questions submitted by the JTEC team prior to their visit. Mr. Kasai stated that only the "management" questions posed by the JTEC panel apply to AIST. [A complete list of questions posed by the panel is included in the Yaskawa Tsukuba Research Laboratory site report, pp. 225-236.]

F.12. How much funding is being directed into MEMS research and development in your organization? What percentage is this amount of the total R&D budget?

AIST does not have the necessary statistics to answer this question. However, AIST estimates that, within the 23 Japanese companies participating in funded research, there are 270 researchers working on micromachines, of which approximately 190 are funded by the Japanese national government.

Based on an informal telephone survey, AIST estimates that a number of large Japanese companies participating in AIST's micromachine program have very minor internal R&D activities in this area -- outside of their normal semiconductor R&D programs which include substantial efforts in silicon micromachining and their contributions to the AIST program. One company estimated its expenditures in the neighborhood of 200 - 300 million per year.

F.20. The first generation of university graduates who specialized in MEMS are now finishing their degrees. Are employment opportunities for these students plentiful in Japan?

Japanese companies hiring students out of university programs are not looking for specific skills (i.e., MEMS fabrication expertise); they hire for lifetime careers, and are looking for bright young people with general technical skills who can be trained to learn whatever is needed for the wide variety of specific projects they may get involved in at the company during their careers.

G.7. Are you involved with joint development with other organizations? What kind (other companies, universities, foundations, etc.)?

AIST can only cite examples from its own experience (see text above); AIST is not aware of all the cooperative efforts of the many companies involved in the micromachine program with the other project. National laboratories funded under AIST also have their own cooperative activities with individual companies, inside and outside the purview of the MITI program.

A.12 Are room-temperature superconductors being explored, and if so, what materials and processing techniques are being used?

The micromachine program is not involved in room temperature superconductivity R&D. However, there are other MITI programs that do fund such research, including some that are involved in research in micromachining of such materials. The Sunshine Project also has some related activities -- e.g., wire fabrication for use in superconducting energy storage coils. Prof. Fujita of the University of Tokyo is also working in this area. [The Atom Technology (i.e., nanofabrication) program is listed in the 1993 AIST brochure. According to Mr. Kasai, the micromachine program and the atom technology program are two of the newest ISTF programs at AIST, and are getting favorable budgetary treatment compared to other programs. Yet the micromachine program has not been fully funded. Mr. Kasai mentioned earlier that loss of tax revenues due to the Japanese recession has directly affected his program's budget.]


Question: Does the power plant application apply to nuclear plants?

Kasai: Yes, in principle, but remember that this is not an actual application -- just a vehicle to stimulate development of basic enabling technologies for future (real) applications.

[Mr. Kasai went on to explain that AIST's mission is to stimulate development of generic technology -- Japanese companies are expected to develop and market real applications after the termination of the micromachine project.]

Question: How are projects evaluated?

Kasai: [There is a] two-tier approach:

1. Management/supervisory (what's appropriate use of taxpayers funds, budgetary analysis, etc.)

2. Technical evaluation (were technical goals achieved?, etc.)

As mentioned before, AIST is now considering a third tier: methodology/metrology evaluation (e.g., how do you measure success in a micron scale device?)

Jacobsen: Who owns the equipment? Can companies use facilities funded by AIST to work on a specific product?

Kasai: All equipment is owned by NEDO. When a project is completed, the equipment is returned to the government. MMC research is only "pre-competitive." Conversely, AIST has asked Nippondenso for use of its airbag sensor technology as a base for development of the MMC project!

Wise: Is technology developed at each company under MMC funding available to all companies participating?

Kasai: "Foreground" rights are 50% owned by NEDO. Participants creating foreground [rights have] the other 50%. Other Japanese companies that pay taxes may also have access; but as non-members, they have to pay 100% usage fee for those foreground [rights].

[Mr. Kasai explained further. The final criteria for evaluating the micromachine program will be as follows: (1) Did it develop useful technologies and concepts? and (2) Management criterion: is it bottom-up technology?]

Holdridge: Is there a way to count resources in other MITI/AIST research programs that are relevant to MEMS?

Answer: Basically no.

Question: Are there formal evaluation meetings?

Answer: Yes. Constant evaluation [is being] carried out by NEDO and AIST. Final decisions on allocation of research funding are made by AIST in consultation with NEDO.

Published: September 1994; WTEC Hyper-Librarian