Keynote Speakers

Martin Riedmiller

Title: Robots that learn from scratch

Abstract: Being able to autonomously learn ‘from scratch’ - i.e. with a minimum amount of prior knowledge - is a key ability of intelligent systems. This credo is the driving motivation behind our research on reinforcement learning methods for the control of dynamical systems. While we have seen tremendous progress in the area of deep reinforcement learning in the last couple of years, its direct application to real systems still remains a challenge. Key requirements for agents mastering the real world are data-efficiency and reliability of learning, since data-collection in real environments, e.g. on real robots, is time intensive and often expensive. I will highlight two main areas of progress that we consider crucial for progress towards this goal - improved off-policy learning methods from large data sets and better exploration. I will give examples of simulated and real robots that, by following these principles, can learn increasingly complex tasks from scratch.

Biography: Martin Riedmiller is a research scientist and team-lead at DeepMind, London. Before joining DeepMind fulltime in spring 2015, he held several professor positions in machine learning and neuro-informatics from 2002 to 2015 at Dortmund, Osnabrück and Freiburg University. From 1998 to 2009 he lead the robot soccer team ‘Brainstormers’ that participated in the internationally renowned RoboCup competitions. As an early proof of the power of neural reinforcement learning techniques, the Brainstormers won the world championships for five times in both simulation and real robot leagues. He has contributed over 20 years in the fields of reinforcement learning, neural networks and learning control systems. He is author and co-author of some early and ground-lying work on efficient and robust supervised learning and reinforcement learning algorithms, including work on one of the first deep reinforcement learning systems.

Koichi Suzumori
Tokyo Institute of Technology

Title: Soft Robotics as E-kagen Science

Abstract: In this presentation, I will discuss three topics on soft robotics.
  1. Since 1986, I have been developing various types of soft actuators; they include pneumatic rubber actuators, thin artificial muscles, functional rubber surfaces. I will be discussing them and their applications to medical robots, soft power support suits, musculo-skeletal robots, and Giacometti robots.
  2. Last year, the MEXT KAKENHI project on soft robots was initiated in Japan with a budget of 1.2 billion yen and a research period of five years. Approximately 20 research groups participate in this project that I will now introduce.
  3. I think soft robotics is a value changer in robotics. Soft robots are considered “bad robots” from the viewpoints of traditional robotics that seek power and accuracy. However, soft robots realize safety, adaptability, and compliance easily, which are important properties in several new robot applications. I will discuss my opinions on the significance of soft robotics with the help of a Japanese word “E-kagen”, which has two contrasting meanings. On the positive side, it could mean suitable, adaptable, and flexible; on the negative side, loose, imprecise, and arbitrary. It is very interesting that these two opposite meanings correspond to the good and poor aspects of soft robots.
Biography: Koichi Suzumori received his Ph.D. degree in mechanical engineering from Yokohama National University in 1990. He worked for Toshiba R&D Center from 1984 to 2001, and for Micromachine Center, Tokyo, from 1999 to 2001. He was then a Professor at Okayama University from 2001 to 2014. Since 2014, he has been a Professor at Tokyo Institute of Technology. He has developed various types of new actuators and applied them to new robots including soft robots, micro robots, and tough robots. He established a start-up venture company, s-muscle Co., Ltd., in 2016, which puts his soft thin artificial muscles into practical uses.