April 1997 GMD Technical Report #1077 Recent Developments in Japanese Robotics Research – Notes of a Japan Tour – Uwe R. Zimmer † , Thomas Christaller † , Christfried Webers ‡ German National Research Institute for Information Technology (GMD) † Schloß Birlinghoven, 53754 Sankt Augustin – Germany ‡ GMD liaison office, 7-5-56 Akasaka, Minato-ku, Tokyo 107 – Japan Robotics appears to be a very lively and fruitful field of re- search in Japan. Some of the research topics cannot be found elsewhere at all, and some are significantly ad- vanced. Discussing this impression, a collection of labora- tories is introduced with their most outstanding topics. Moreover some background information about research plans, politics, and organisations are given. This report is based on the notes of a trip through Japanese laboratories in February 1997. 1. Motivation and Criteria The robotics research scene in Japan is a widespread and lively community. From this scene, research top- ics and projects that are unique or original will be highlighted. This report is not meant to be a strict sci- entific discussion of the subjects, but to set a spotlight on recent trends and developments in the Japanese robotics research community and to give links for deeper investigation. Criteria used for the description of groups and projects include: • Original and outstanding research topics. • Basic research issues. • Disciplines represented in the group, thus the interdisciplinarity of the group – a key issue for most parts of robotics research. • International exchange (visiting researchers). • Intended or implemented applications. • Industrial relationships. • Participation in research funding programs. • Organisation or initiation of conferences, work- shops, and scientific organisations. • Special aspects/notes. The discussed groups and projects are a selection of the laboratories known or have been visited by the authors only. Thus this report focuses on some recent work instead of listing the full range of robotics groups. Up to now, there are more than 150 Japanese research groups concerned in robotics research or closely related topics [73]. And even in this listing the ‘classical’ control theory oriented stream of robotics, mainly relating to manufacturing manipulators, is not covered completely. The description of individual laboratories is organ- ized in three sections: • Biological orientation: The motivation to under- stand biological creatures and to mimic certain ob- served aspects is a central issue in these laboratories. Biologically plausible solutions are usually preferred over technologies, exploiting global observers or high speed communications. • Learning & cognition: Here the understanding and design of artificial creatures is the central as- pect. Especially adaptability to current working environments as well as the understanding and modelling of complex interactions between robots themselves and with their habitat is discussed. • Mechatronics: By studying the complex possibili- ties of kinematics with many degrees of freedom and understanding high speed feedback systems, a third important aspect of robotics research is ad- dressed. These categories are anything but rigid. The selection being done here is based on the main motivation of each group. Obviously, solutions originating from investigations in many degrees of freedom kinemat- ics can also be biologically plausible. The authors are aware of the fact that the selection of ‘interesting’ topics is to a certain degree related to the research focus and background of themselves. Thus the robotics research activities of the authors should be mentioned briefly by keywords: Autonomous sys- tems [10], [64] (especially mobile systems in natural environments), learning and adaptation [9], [21], [63] (on all levels of control/cognition), complete systems [11] (i.e. ‘closed loop’ systems), and complex behav- iours [12], [65] (together with fruitful experimental setups evaluating these behaviours).