Computer architecture for intelligent robots

Yutaka Kanayama,Shin'Ichi Yuta
DOI: https://doi.org/10.1002/rob.4620020303
1985-01-01
Journal of Robotic Systems
Abstract:Abstract This article proposes a computer architecture suitable for intelligent robots, especially for self‐contained intelligent mobile robots. The main principles proposed by the authors are: (1) The robot should be a multiprocessor system with a master, several slave modules and a console. A simple star connection is employed. (2) The master carries user's programs written in a high level language with which a programmer is able to use all basic functions in the robots. It should have a special purpose operating system. (3) Each module is an independent microcomputer system loosely coupled to the master and dedicated to an elementary function such as manipulation, locomotion, sensing, or planning. (4) A serial TTL level or RS232C interface is employed between the master and each module. Two self‐contained robots, Yamabico 9 and 10, constructed under these design principles have demonstrated the effectiveness of this proposed architecture.
What problem does this paper attempt to address?