COLUMN ONE : Big Hopes for Tiny Machines : Microrobots and other miniature devices could perform remarkable feats. Americans pioneered the technology, but Japan is gearing up to develop it.
Isemi Igarashi is nursing a beer in a cramped Tokyo sushi bar and expounding on the future for microscopic robots when he notices a mosquito circling his raw tuna.
“Take the mosquito, it’s a fabulous mechanism,” says Igarashi, executive vice president of Toyota Motor’s central research and development laboratory. “It has little sensors that seek out a blood vessel. It makes a cut in the skin with the saw at the tip of its beak and sucks out a precise quantity of blood.
“Don’t you hate shots? A machine built like that (mosquito) could take your blood, diagnose it, and you wouldn’t even notice it.”
Looking further ahead, Igarashi suggests, a microscopic capsule could coast through a cancer patient’s bloodstream on search-and-destroy missions against diseased cells, reminiscent of the movie “The Fantastic Voyage.”
All this may sound fanciful. But Igarashi is part of a small but growing group of visionaries around the world who believe micromachines--miniature machines that can be smaller than a grain of sand--ultimately will have an impact on everything from space exploration to pest eradication.
Finger-size rockets packed with microscopic instrumentation could be sent off into space at a fraction of the cost of current missions. The deep sea could be explored by toylike submarines more capable than larger vessels of withstanding the enormous pressure thousands of feet underwater. Microscopic factories with all but invisible assembly lines could combine single cells into new forms capable of combatting disease in humans or plants.
Tiny robots also could attach themselves to thieves and help police track down the culprits. Or they could inspect cooling pipes of nuclear power plants for cracks.
Those tools may be just the beginning. In a recently published book, Iwao Fujimasa, a Tokyo University professor of medicine, argues that micromachines are at the beginning of the same evolutionary process that created human beings from simple organisms. Eventually, he suggests, machines will be capable of doing much of what humans can do without the shortcomings.
Seeing the potential, Japan has decided to make “microrobots” the nation’s next technology target. Although commercialization is seen as at least a decade away, the Ministry of International Trade and Industry, Japan Inc.’s top strategic planner, plans to plow $200 million into a 10-year project beginning next fall to bring university, government and private industry laboratories together to build first-generation prototypes of these bug-size machines.
MITI’s prototypes will probably be far clumsier than mosquitoes, more like tiny tools than sophisticated robots. Under consideration for development is a catheter tipped with optical sensors and microprocessors to guide it through the human digestive tract. Also possible are drills to cut away plaque in plugged arteries.
The nation’s researchers, sensing a new boom sector, are taking up the banner and rushing to take the lead on this new technological frontier. The new field’s most important annual conference will be held in Japan for the first time starting Jan. 30.
Japan’s cooperative effort is causing concern among U.S. scientists who fear that yet another technology pioneered in the United States is about to be overtaken by Japanese industry. The United States still has the lead, but researchers estimate that U.S. spending in micromachines is not much more than $5 million a year, with most of the money coming from the National Science Foundation.
“If American companies don’t take the right steps soon, we are going to be buying this technology from Japan,” says Richard Muller, a leading pioneer in the field and director of the Sensor and Actuator Center at UC Berkeley. “Japanese companies are willing to gamble on things farther out.”
Researchers have long sought to make machines smaller so they will take less space, use less energy and be capable of taking on more delicate tasks. But there is a limit to how small you can manufacture metal parts. And, like the delicate mechanisms of a watch, the smaller the parts, the harder they are to produce and handle.
All that changed in 1988. Then, Muller’s research group--using photolithography and other techniques developed for semiconductor manufacturing as part of a process called “micromachining”--etched into silicon the gears of a primitive motor 0.0001 of an inch in diameter. That’s about twice the width of a human hair.
The fortunate coincidence of the motor’s being made from silicon means that dozens of them could be carved out of a single silicon chip and combined with semiconductor circuits and sensors. Industry has already learned to combine microprocessors (the brains) and sensors on single chips of silicon.
“Now we can add the arms and the legs,” says Hiroyuki Fujita, an associate professor at the University of Tokyo who is widely recognized as Japan’s leader in the field.
Taking advantage of advances in semiconductor production, potentially thousands of these tiny machines could be manufactured on a single silicon wafer 6 inches in diameter and sawed into little pieces to become the bodies of future robots. Mass produced, they could become so cheap they could be used for menial tasks such as killing pests, cutting grass and pulverizing concrete.
Micromachines are not far from finding applications in simpler mechanisms for valves and mechanical switches in optical fiber cables. Toyota’s Igarashi, who succeeded in combining a pressure sensor with semiconductor circuits more than a decade ago, sees micromachines helping to trouble-shoot problems in auto engines not far in the future.
But the most dramatic applications will be decades away. Researchers have yet to develop microscopic motors that run for more than a few minutes. And, so far, the tiny motors would have difficulty budging a paper clip.
The United States is still far ahead in micromachining research. Small U.S. venture companies such as Novasensor of Fremont, Calif., and IC Sensors of Milpitas, Calif., are using micromachining techniques to build “smart” sensors that can be used to set off air bags in automobiles or automatically adjust a car’s suspension system based on road conditions.
Researchers at such places as Massachusetts Institute of Technology and the University of Michigan continue to come out with breakthroughs far beyond anything done in Japan.
But although the United States, like a well-tuned sports car, is getting a quick start on the new technology, Japan, moving more like a locomotive, may do better over the long haul. Japan starts up slowly as it develops a consensus. But once it is moving, it has the power to pull behind it an entire industrial infrastructure.
Already, Japan’s largest companies are showing far more interest in the technology than any U.S. company. At the international microrobotics conference in Nara this month, several dozen U.S. professors will be attending but only a handful of U.S. companies are represented. By comparison, two-thirds of the 250 Japanese participants will come from Japanese companies.
The broad range of research at Japanese companies gives them an edge over university-centered U.S. efforts.
American Telephone & Telegraph, which had one of the world’s most advanced micromachining research programs, has chosen to pull out because of the technology’s limited potential applications in communications.
By comparison, Japan’s NEC Corp., which manufactures a wide range of products in electrical and machinery fields, is confident that it will find applications.
“There are a whole range of areas in which we can apply this technology,” says Michiyuki Uenohara, former head of NEC’s research efforts and now executive adviser.
And Japanese companies, with strengths in robotics and semiconductors, see the micromachining technology as a natural extension of their work.
Seiko Instruments, long interested in machinery to handle tiny components for its watch production lines, boasts that its new robot arm capable of grasping things 0.3 millimeters thick is an advance in microrobotics. Omron Corp. announced last month a 9-millimeter electrostatic motor three times as powerful as existing motors that size. Seiko needs more delicate machines in the manufacture of the tiny parts for its printers and watches.
But it requires a quantum jump in technology to begin producing components for the micro world. And Japanese companies admit that they are reluctant to invest in this fantasy world of insect robots without government help.
When the announcement by UC Berkeley’s Muller was made in 1988, it was clear that the technology was a strategic one. But few universities had the facilities to do the research. Fujita’s laboratory, probably the best-equipped university lab in Japan for micromachine research, is a small shack off the courtyard of the University of Tokyo’s Production Technology Research Center in downtown Tokyo.
“We needed to find a way to get companies to invest in this research,” says Kimiharu Sato, manager of the Japan Industrial Robot Assn.’s technical department.
The association established a committee made up of a who’s who of Japanese industry to produce a 211-page study on microrobots, complete with illustrations.
The study, completed in March, 1990, and paid for with gambling money from MITI’s bicycle-racing fund, is part science, part fantasy.
To help persuade the Ministry of Finance and the public that microrobots are a worthy cause in which to invest, the study emphasizes the potential savings in medical costs at a time when an aging population is likely to draw heavily on social resources.
“We could do surgery at home, you could save the costs of hospitalization and surgery,” says Kenzo Inagaki, deputy director of MITI’s Industrial Machinery Division.
The technology would also be user-friendly. While people might worry about bringing a human-sized robot into their home for fear it might go crazy and hurt someone, the report says, a miniature robot would not elicit such fears. “If something goes wrong with one of these,” says a Japanese industry report, “you could swat it with a slipper and kill it.”
The study also addresses the question of production. Microrobot factories, for example, would have to be fully automated because no humans could handle the tiny machines.
The report, which represents a compromise between electronics-oriented members of the committee and its machinery participants, recommends that MITI put together a project to build two robots, one microrobot for medical applications and a slightly larger miniature robot for factory maintenance. Those two fields--medical applications and factory maintenance--represent a $300-billion market in Japan, according to Sato.
Last autumn, the robot association study landed on a desk at MITI’s Large Scale Projects division along with about 40 other proposals--everything from space robots to biocomputers (computers that use living organisms, such as bacteria, for sensory or informational purposes). In August, MITI announced that it had chosen micromachines as its next project.
“It is a good project because it will have a positive impact on human welfare as well as being good for industry,” says Masahiko Kobayashi, who helped make the decision as deputy director for Large Scale Projects at MITI.
Because commercialization is at least 10 years away, it will be easier for companies to cooperate with each other without worrying about competition. “It is a marathon race in which only the first mile has been run,” he says.
MITI is now looking at research proposals. A group in MITI’s electrotechnical lab is looking into ways to have teams of such machines working in tandem, much as numerous muscles are used to control the movement of a single finger. Thousands of microscopic devices embedded in an airplane wing could, working together, twist the airplane wing, offering far greater maneuverability.
MITI must also decide whether to create a joint laboratory as most researchers favor, or to contract out research, as administrators of existing laboratories prefer. And university researchers say MITI is too timid in setting goals for the project.
“They aren’t proposing to do enough basic research,” says Fujimasa of Tokyo University. “We are going to be criticized (from overseas) for not contributing enough new to the field.” Researchers also complain that companies, not universities, will be the primary beneficiaries of the project.
While the argument continues, the project is already helping to generate excitement. “MITI’s project lends legitimacy to the field,” says Fujita.
Young researchers at Japan’s corporate laboratories, many of whom were conducting “under the table” research on micromachines, can now lobby for corporate funding.
Kazuo Ozawa, manager of the central research laboratory at Omron Tateishi, a manufacturer of computerized control equipment, says much of its current effort involves information gathering. “We want to tie up with the best labs in the world and absorb that information quickly,” says Ozawa. However, he says, the MITI project could give him enough clout to get the company to build a separate laboratory.
Local governments around Japan are scrambling to establish themselves as micromachining centers in hopes of attracting future factories. Nagoya sponsored an international conference in the city in October and is talking about putting up $77 million to establish a micromachine laboratory to attract researchers.
Nagoya is already a microrobot center of sorts. Toyota, one of the leading research companies in the field, is located nearby. And Nagoya University professor Toshio Fukuda came out with a tiny robot that can make its way through networks of pipes. At 6 millimeters in diameter, the robot is a giant in the micro world, but it crawls through the pipe powered by an external magnetic pulse, a system Fukuda hopes can be used to power microrobots. The robot could be used inside cooling pipes of nuclear power plants to inspect for cracks, for example.
But even regions without expertise in the area see a future in microrobots. The mountain prefecture of Yamagata does not have a single scientist in the area doing work on micromachines. But, to get an early start on what it sees as an emerging field, the local government commissioned Shigetoshi Oshima, a specialist in superconductors at Yamagata University, to write a 250-page report on the technology’s potential.
The prefecture is now trying to raise several million dollars to establish a laboratory in the region where corporate and university researchers can pursue micromachine-related research. “We want to raise the technology level of Yamagata in this field,” Oshima says.
Times staff writer Carla Lazzareschi in Los Angeles contributed to this story.
A MICROMACHINE PROTOTYPE
* This miniature manipulator, a prototype to be developed by Japan’s Ministry of International Trade and Industry, will have tiny hands, scalpels and drills that can be used for surgery in the intestines and stomach.
Only one-fifth of an inch thick, the entire manipulator can be passed down the throat and into the digestive system.
* Unlike current systems that are inflexible and painful, this manipulator uses its own brains and sensor to change its shape to match the passageway it is going through. Current systems can merely take pictures and do minor surgery in the stomach.
* The manipulator would be controlled by a surgeon operating from a terminal. With any luck, says a MITI official, the system would also be capable of doing surgery in the liver and the pancreas.
