Exploring intelligent movement through the machinery of the body

In interdisciplinary research, anything can happen. That’s what makes it so fascinating.

Hosoda, Koh

Graduate School of Engineering Science
Adaptive Robotics Group

Hosoda, Koh

Professor Hosoda investigates the “machinery”, in other words the movement systems and functions, of humans and other living things using robotic replicas of their musculoskeletal structure. This investigation is now branching out to look at how this machinery is used by the brain to control movement.

A jumping robot that uses the machinery of the human body

In order to study the machinery of the body, I make robots modeled on humans’ lower body structure using artificial muscle tissue that responds to air movement. These robots make effective use of bodily machinery in their movement, reducing the number of computations needed to control them and making operation possible with just a small computer and a few sensors. It is thought that in humans, too, rapid muscle movements such as jumping cannot be controlled simply by signals sent from the brain.

hosoda2

Delving deep into muscular function

Behind our shinbone, for example, there are two muscles: the soleus and gastrocnemius. We know that the soleus moves the ankle joint and the gastrocnemius can move both the ankle and the knee simultaneously. But at this stage, it is not known exactly what functions each of these muscles possesses. Both of them can move the ankle, but we are not sure exactly how their roles in this movement are divided.
All over the human body, there are similar examples of multiple muscles moving the same joint simultaneously. It is thought that these muscles help control body movement during physical activity. Approaching this phenomenon using robots enables us to conduct experiments on muscle function that would be impossible using real humans; halting the activity of a certain muscle, for example.

Disciplinary differences also mean differences in language and perspective

My laboratory has been pursuing joint research with experts in biomechanics, biology, and medicine since long before the Humanware Innovation Program (HWIP) began. When HWIP was launched, I didn’t set out on new interdisciplinary research; I just continued with the research I was already doing.
I find that people working in different disciplines use different vocabulary and different ideas, but as I talk with them, I often discover new ways of thinking about things. Even if I have acquired knowledge of the area in question from books, I am still impressed by the differences in approach once the interdisciplinary research actually gets under way.
When I was first starting out on my current research, I remember showing one of our humanoid robots to a researcher in a different field and being told: “this is not at all the same as a real human.” When I got the researcher to explain exactly what he thought the differences were, I realized that he looked at things in a totally different way. This experience was an immediate inspiration to me. I find it really rewarding when such discoveries lead to advancements in research.

Cultivate the “gems” obtained through dialogue with others

hosoda3I believe that interdisciplinary approaches evolve naturally in the course of research. They are not something that can be designed in advance. It is this sense of the unexpected that makes research so enjoyable. Interdisciplinary research often begins with something that stands out to you when you listen to what experts in other fields have to say—a small “gem” in your conversation, as it were. When students from other fields visit my lab as part of the HWIP lab rotations, I didn’t go out of my way to explain what we do as “interdisciplinary research”. I just tell them what our research was about. I hope that they will absorb what I say into the knowledge and expertise they already have, and make connections with the interests and challenge they face in their own research.