WASEDA ONLINE

RSS

The Japan News by The Yomiuri Shimbun

Home > Research > Knowledge Co-Creation - Profiles of researchers

Research

Knowledge Co-Creation - Profiles of researchers

Pursuit of Human Mimetic Technology beyond Humanoids

Hiroyasu Iwata
Associate Professor, Waseda Institute for Advanced Study

Development of the world's first robot for cracking an egg

When I was a fourth grader, I listened to the theory of necessary cosmic velocity for a rocket to get out of the gravity zone in a scientific television program, and was astonished purely by "the beauty of the motion of an actual object as expected in accurate theory." Since then, I have yearned for the scientific technology for space rockets, and I entered the School of Science and Engineering in university.

WENDY, the robot with "flexible fingers" that cracked an egg for the first time in the world

When I was a 3rd year undergraduate student, I applied for the enrollment in the fluid engineering laboratory, which is closely related to space science and technology as the first choice. However, when I dropped at the laboratory of Professor Shigeki Sugano (present: Professor of Faculty of Science and Engineering, Waseda University) in robot engineering, I saw a "flexible robot arm" and was astonished by the uniqueness of the idea and the outstanding technology. Although the deadline for application was over, I patiently requested the university to change my laboratory choice to Professor Sugano's laboratory. As a result, my request was accepted, although I had to write an essay repenting my mistake with one page of A4 paper. This became a turning point.

At that time, Sugano's laboratory engaged in the research and development for producing "a robot that can coexist with human beings" for a super-aging society, and so I could do my graduate research in the national large-scale project. Through trial and error, I succeeded in developing the life-size dual-arm manipulator by installing a structurally-flexible mechanism in the joints. Just after that, when I was a 1st-year student in the master's program of the Graduate School, I was fortunately given an opportunity to participate in the WENDY project for the development of robots that can coexist with human beings, and succeeded in developing the technology for robot hands that can crack an egg for the first time in the world.

"Human Mimetic Technology" I aim for

After completing the doctoral program, I was appointed as the project leader of TWENDY-ONE for the development of robots that can coexist with human beings. This is a 5-year project aimed at actualizing a full-scale "useful robot" that assist us in nursing care and housekeeping. My role was to create epoch-making ideas and wisdom and embody them in close cooperation with students and engineers in the private sector, to proceed with the entire development project. I aimed to embody a human mimetic robot that supersedes conventional humanoid robots.

Based on WENDY (Left in the photo), TWENDY-ONE, the robot that can coexist with human beings, (Right in the photo) was developed.

The conventional WENDY was mounted with human-like arms and fingers, but their functions were far behind the pliability of human arms. In this circumstance, we initiated the project with a determination to formulate a theory for designing, from scratch, a robot that possesses necessary functions to coexist with human beings, safety, motion flexibility, communication capacity, and design, etc. and with a strong belief that by formulating an accurate theory, we will be able to actualize a robot that can coexist with us.

In robot development, it is difficult to densely assemble various devices and components in limited space and orchestrate all of them while activating every joint without causing disconnection or mechanical interference. Nevertheless, TWENDY-ONE also attempted to install special skin materials, a joint mechanism, and sensors in the hands and arms, and then achieved both the preciseness in motion and the harmlessness to human beings, which had been considered the most difficult in designing a robot that can coexist with human beings, at high levels.

TWENDY-ONE assisting in housekeeping, using tools skillfully

After announced in Nov. 2007, the news was transmitted to around the world, attracting public attention. In 2007, TWENDY-ONE has attained the METI's goal of developing robots by 2017. At that time, most of developed countries felt the necessity of nursing-care robots to cope with the upcoming super-aging society, but they could not embody such robots. In that situation, our robot was released, and so TWENDY-ONE attracted significant attention from the media and robot experts around the world, as a practical solution.

For 5 years, I researched all night about 3 times a week, and made discussions for 20 hours a day, to pursue production without making compromises. This experience significantly contributed to my career.

For the establishment of system integration engineering

During the TWENDY-ONE project, I felt the necessity to acquire a special skill that can be dubbed "system integrator," which determines whether or not the project is successful. When I think about the subject "system integration engineering," which is still to be formulated, I consider that the following three are important.

First is "the design of relation with users." We should pursue good relations between human beings and robots, and put them into practice. For example, what is "nursing care"? Does it mean to hold, heave, and convey a bedridden person? We rather thought it is to assist them in getting up and standing up by themselves.

The nursing care of TWENDY-ONE is "assistance in independence" or "minimum support" for independence. As a bedridden person becomes able to move more by his/herself, the degree of assistance is lowered gradually. Actually, over half of elderly people who require nursing care are relatively healthy with a care need level being 2 or less, and so it is considered unnecessary to fully assist all of them. It is important to design the robot from such a social perspective.

Second is "the comprehensive design of hardware and software." This means to design the robot's body and intelligence in a balanced manner. For example, a thin acrylic plate cannot be picked up by the fingers without the nails. It is very important to have a comprehensive perspective of designing a system by combining the base hardware called the body and the software called artificial intelligence to manipulate the body in a balanced manner, that is, designing the shape, functions, and mechanisms concurrently.

The roles of the system integrator for a large-scale project are diversified, but the success of the project depends on whether we can find clues to solving problems.

The essence of the hardware design for TWENDY-ONE is "the softness of machines." By adding softness to the skin and joints of the arms and trunk, it is possible to protect people from the shock of the unexpected conflict with the robot. When softness is added to the fingertips, palms, and finger joints, the contact surface between the robot skin and an object to pick up enlarges and stability increases even if the object has a complex shape, which improves dexterity. By focusing on the design for actualizing such softness, we succeeded in enabling a robot to crack an egg and handle a thin straw with simple control, for the first time in the world.

Third is "the balancing of functions to develop a trade-off relation." This means to solve discrepancies among elemental technologies when integrating various elemental technologies. This is indispensable for actualizing many functions at the same time like TWENDY-ONE. Every time I was confronted with a trade-off problem-sophistication vs cost reduction, flexibility improvement vs downsizing, and functionality vs design, I cudgeled my brains and solved it, utilizing the system integrator (me). System integration fails, if we cannot achieve a good balance.

Nursing-care robot TWENDY-ONE to assist people in independence at minimum

System integration engineering can be said to be the integration of all engineering studies. When negotiating with a specialist in each field for the cooperation with companies, etc., the specialist is more versed in technology in some cases, but we must express our requests for attaining our goals undauntedly. To do so, it is important to develop engineering arguments, find original philosophy fusing academic knowledge of human science, phenomenology, sociology, and theory of business administration, etc., and prepare persuasive stories and visions. Accordingly, we need to have a good sense of balance so as to seek satisfactory solutions to attain goals while simulating various cases under every possible condition and taking into account actual constraints.

I want to verify the plasticity of the brain

For the past several years, I have engaged, with interest, in the development of a new rehabilitation system that combined cognitive behavior therapy and robot technology (RT), which were originally developed for assisting those who are suffering from hemiplegia due to cerebral infarction, etc. Conventionally, rehabilitation was carried out mainly with exercise therapy for strengthening muscle, and cognitive therapy for recovering brain functions was not conducted so often. Meanwhile, cognitive behavior therapy is aimed at recovering functions by stimulating the brain, because the body is normal and the function to transmit signals from the brain to the body malfunctions.

In the case of hemiplegia, the signals from the paralyzed body part do not reach the brain. The therapists of cognitive behavior therapy stimulate the paralyzed body part with their hands, to make the patients concentrate on body sensation, but this is difficult and the therapy is unsuccessful in many cases. With a system based on biofeedback, patients would be able to rehabilitate themselves. Then, I developed the RT for supporting sensation by detecting the pressure on the paralyzed sole with sensors and conducting biofeedback to normal body parts (international patent is pending).

Due to hemiplegia, the paralyzed foot does not step on the ground normally, and weight loading is biased to the outer side, causing clubfoot deformity, in many cases. To cope with it, oscillating motors, etc. are mounted on the normal body parts, such as the back, and if the foot steps normally, the "OK" signal is transmitted, and if the foot does not step normally, the "NO" signal is transmitted through tactile stimulation.

With the biofeedback technology, it is possible to provide bypass routes that virtually act for broken nerves. Actually, the effect of improving clubfoot was observed in the empirical experiment targeted at patients. In addition, through the experiment based on brain neuroscience, it was found that the paralyzed side sensory area is activated when rehabilitation is conducted based on the stimuli of feedback. This result reinforces the possibility of enhancing the plasticity of the brain by conducting rehabilitation with the RT that makes patients concentrate on body sensation.

With the biofeedback technology, it is possible to provide bypass routes that virtually act for broken nerves. Actually, the effect of improving clubfoot was observed in the empirical experiment targeted at patients. In addition, through the experiment based on brain neuroscience, it was found that the paralyzed side sensory area is activated when rehabilitation is conducted based on the stimuli of feedback. This result reinforces the possibility of enhancing the plasticity of the brain by conducting rehabilitation with the RT that makes patients concentrate on body sensation.

One example of the RT for complementing sensation. In this case, I adopted a biofeedback system that detects the pressure on the paralyzed sole with sensors and stimulates the normal limb with pressure based on the detected pressure.

Hiroyasu Iwata
Associate Professor, Waseda Institute for Advanced Study

Earned a doctoral degree in engineering at Graduate School of Science and Engineering, Waseda University (intelligent machinery and machine system). Served as assistant of Department. of Mechanical Engineering, School of Science and Engineering, Waseda University; lecturer of Graduate School of Science and Engineering, Waseda University; lecturer of Consolidated Research Institute for Advanced Science and Medical Care, Waseda University (ASMeW); associate professor of ASMeW; and now serve as associate professor of Waseda Institute for Advanced Study (tenure-track program) from Oct. 2007. From 2003 to 2007, served as the project leader and system integrator of the TWENDY-ONE project http://twendyone.com/index.html. Received the young scientists' prize from the Minister of Education, Culture, Sports, Science and Technology in 2009. (http://www3.atword.jp/wasedatech/archives/531).

WASEDAWASEDA UNVERSITY Research Promotion  http://www.waseda.jp/rps/