I would like to create the “robot” as a new kind of living being.


Dylan F. Glas | Research Engineer, Department of Network Robots, Intelligent Robotics and Communication Laboratories
Born in the U.S.A.
Master’s degree: Aerospace Engineering, Massachusetts Institute of Technology.
Main areas of research are teleoperation for social robots, network robot systems for human-machine nteraction, ubiquitous sensing, artificial intelligence, etc.

Rather than work which harms people, I want to do work which supports people.

I was born in the United States, and I received my master’s degree in “Aerospace Engineering” from MIT.
After that, I worked as an IT consultant, taught English at a public high school in Japan, and participated in a project in which I went to Jerusalem and taught programming skills to Israeli and Palestinian high school students. As these members of the younger Middle Eastern generation continue their studies and careers in technology, I strongly hope that the relationships they have built through the program will be a catalyst for positive social change and help to improve the situation in that conflicted region.

Although there are many opportunities to pursue robot-related work in the United States, much of that research is supported by the military. Although the availability of funding is lucrative, the results of that work are ultimately used to harm people. Particularly after my experiences in the Middle East, I did not want to engage in such work. However, in Japan, I found that peaceful robots were being developed for everyday life, which could hug children and carry groceries for elderly people. I was quite impressed with their way of thinking, and this is why I am here at ATR now.

Making robots which can be perceived as lifelike social peers.

In my work, I have explored a variety of research themes.
In past work, I developed a system which can perform high-precision tracking of pedestrian positions using laser range finders, and a system which enables a single human operator to support teams of up to four social robots simultaneously.
In my current research, I am developing techniques using machine learning to generate more humanlike social behaviors for robots. It is quite challenging for a robot to learn to perform natural social interactions autonomously. For example, when greeting someone, a human might say something appropriate to the current situation, such as “It’s warm today, isn’t it?”, “Who is this? Your granddaughter?”, or “You seem quite relaxed today”. This is very difficult for a robot to do – responding smoothly to environmental situations and personal interaction history. It is possible to explicitly program many rules to create behaviors like this, but I think it is more robust and scalable to collect examples of what real people would say in those situations, and have the robot learn the behavior for itself.

A robot needs to judge every situation and respond appropriately. This is true for explicit actions like utterances as well as simpler, implicit actions. For example, one of our robots reacts when a person touches its body by quickly turning its head to glance at the place where it was touched. This reactivity helps the robot to be perceived as more lifelike. Such behaviors cause people to perceive the robot as a social peer, and not just a device.
My goal is to build robots that seem genuinely alive, even to the engineers who built them and understand how they work. Robots are fundamentally different from people, and I do not want a robot to exist simply as a copy of a human being. Instead, I want to establish robots as a new category of intelligent living being, distinct from humans and able to be seen as social peers in their own unique way.

Never Stop Learning and Never Stop Teaching. It is my principle.

Why do I want to work with robots? There are two reasons.
The first is that I want to improve society by developing robots which help and support people. Technology often serves as a dividing and dehumanizing force in our lives, but I would like to use robots to make our lives more rewarding and to make our social interactions with each other more warm and meaningful.
The other reason is that the dream of creating a robot as a living artificial being gives me inspiration and motivation. Inspiration is one of the most valuable and rare resources in our world today. Inspiration is absolutely essential in our lives – it fuels innovation, motivates people to create and achieve, and fills our lives with meaning and enjoyment. For me, robotics is a great source of inspiration – it represents one of the exciting frontiers of engineering and science. This is why I love to work with robots.
Finally, it is my belief that the process of brainstorming, of creative synthesis through sharing and contrasting ideas with others, is one of the most valuable and rewarding ways we can interact. “Never Stop Learning and Never Stop Teaching” is my principle. It is very important for all of us to search out and listen to new ideas, and to share our knowledge with others.
Of course there are language and cultural barriers, but I would like to continue development of robots here in Japan. There is an abundance of wonderful talented people and cutting-edge technology in Japan. Therefore, I am excited to know that there is a special place for people to gather and exchange their grand ideas here in Osaka.