Masatoshi Ishikawa, Dr. Eng.
Professor of Department of Information Physics and Computing of the University of Tokyo
A robot developed by Dr. Masatoshi Ishikawa, a professor of the University of Tokyo, wins the rock-paper-scissors ("Janken") game against a human 100% of the time. The potential of this "intelligent robot" capable of moving quickly using a high-speed vision sensor was transmitted from Japan all over the world. The sensing technology and image processing technology enabling the Janken robot to see have a great potential for a variety of industrial applications. Dr. Ishikawa has been involved with industry-university collaborations using university's knowledge and designing a system to create ventures. We talked to him about the future created by scientific technology and the potential of Tokyo.
Q: Your Janken robot can detect human hand movements with a high-speed vision sensor and win the game every single time without cheating. Please tell us about your research theme you are working on right now based on the robot.
A: The Janken robot throws rock, paper, or scissors in less than one-thirtieth of a second after the human player completes its move. Since human eyes can process images at 30 frames per second, the player can’t comprehend that the robot cheated and the robot wins the game 100% of the time. We are aiming to develop and diffuse high-speed vision capable of recognizing and processing images at much higher speed than human eyes just like this Janken robot.
Q: What and how will things change with high-speed vision?
A: High-speed vision will help us create robots whose senses and work speed go far beyond human abilities. With the birth of high-speed robots, labor-intensive work by humans will be increasingly replaced by those robots. Therefore, the way industries value low labor costs will change greatly. Moreover, I believe that workers who currently engage in labor-intensive work will engage in knowledge-intensive work such as designing, maintenance, and monitoring of the robots.
Q: You are closely associated with venture companies. Do you think Tokyo offers a rich environment to venture businesses?
A: Tokyo has a large group of people with different ideas so it is definitely a favorable place to start a venture business. There are a variety of facilities and places to meet people, but without any facilitators or directors who can act as intermediary, it is hard to boost venture creation. So, I believe that developing such human resources is our future challenge that needs to be addressed.
Q: You insisted the importance of looking at underlying needs in society when commercializing science technology. Please tell us more about your idea.
A: Proof of Concept (POC) is a term referring to evidence, which establishes that a new concept or theory is feasible. POC is essential for identifying underlying needs. The process of POC is necessary when someone wants to realize a new method or idea, not to imitate something or solve problems. Since this process entails some risks, a social structure that hedges against those risks becomes important. For instance, a system to aid collaboration between venture companies and universities will hedge against risks of venture companies.
Q: What is your opinion about AI that controls robot’s intelligence?
A: Dr. Rodney Brooks, founder of iRobot who created a cleaning robot Roomba, explained two types of AI on his paper titled Elephants Don’t Play Chess. Until his paper was published, the mainstream logic of AI research was to build AI that can beat a chess champion. On the contrary, elephants don’t play chess, but they are intelligent enough to communicate with humans. There is AI that can create existence like elephants. What elephants represent in the real world is, for instance, AI for autonomous driving. Our team is working on this type of AI and we call it intelligent system or behavioral recognition system to differentiate this AI from the other one. This is the area of expertise for Japanese researchers. Our team is moving forward with the research and development as we realize that our AI research has great potential that may lead to the Fourth Industrial Revolution*1 and Society 5.0*2.
Q: And finally, as a researcher of robots and high-speed vision sensor, what is your opinion about conducting research in Tokyo?
A: I believe that Tokyo will become a global leader of this field based on its collective strength including cutting-edge image sensors, robots, and processing technologies. My goal is to network sensors, cameras, and displays together and conduct verification experiments on autonomous driving and disaster/crime prevention to expand a safe “Smart society” in Tokyo. I’m sure that this effort will help build infrastructure for the Tokyo 2020 Olympics Games.
*1 The Fourth Industrial Revolution encompasses the Internet of Things meaning that every object is connected to the Internet and communication is enabled by sensors installed on the objects. This allows networking between factories and connection of various business resources, such as objects, services, and data. This concept, has quickly spreadfrom the German government’s industrial policy Industrie 4.0, has quickly spread around the world.
*2 Society 5.0 is one of the pillars of the 5th Science and Technology Basic Plan approved by the cabinet in January 2017. This concept envisions a future society, or ultra-smart society, where cyber space and the real world are highly integrated by taking full advantage of ICT. In this regard, it is clearly different from the Industrie 4.0 focusing on smart manufacturing.Reference: