Learning-based automatic generation of collision avoidance algorithms for multiple autonomous mobile robots

Yukiyoshi Fujita, Satoshi Fujita, Masafumi Yamashita, Ichiro Suzuki, Hajime Asama

Research output: Contribution to conferencePaper

2 Citations (Scopus)

Abstract

Using a model based on the omni-directional robots developed at the Institute of Physical and Chemical Research (RIKEN), we discuss the possibility of automatically generating a collision avoidance algorithm for autonomous mobile robots. To this end, we show that an effective collision avoidance algorithm for two robots can be generated by a very simple learning algorithm that simulates a naive human trial-and-error learning process, using only the robots' sensor outputs and a suitable reward function, where the exact form of the reward function is also learned autonomously by the robots. We also discuss how a robot can use its `experience' gained in a simple environment to adjust itself to a more complex environment, by automatically generating a collision avoidance algorithm for a three-robot situation utilizing a reduced state space resulting from the learning process for the case of two robots. The results of computer simulation and the experiments conducted at RIKEN using physical robots demonstrate the effectiveness of the collision avoidance algorithms generated and our learning-based approach.

Original languageEnglish
Pages1553-1558
Number of pages6
Publication statusPublished - Dec 1 1998
EventProceedings of the 1998 IEEE/RSJ International Conference on Intelligent Robots and Systems. Part 1 (of 3) - Victoria, Can
Duration: Oct 13 1998Oct 17 1998

Other

OtherProceedings of the 1998 IEEE/RSJ International Conference on Intelligent Robots and Systems. Part 1 (of 3)
CityVictoria, Can
Period10/13/9810/17/98

Fingerprint

Collision avoidance
Mobile robots
Robots
Learning algorithms
Sensors
Computer simulation

All Science Journal Classification (ASJC) codes

  • Control and Systems Engineering
  • Software
  • Computer Vision and Pattern Recognition
  • Computer Science Applications

Cite this

Fujita, Y., Fujita, S., Yamashita, M., Suzuki, I., & Asama, H. (1998). Learning-based automatic generation of collision avoidance algorithms for multiple autonomous mobile robots. 1553-1558. Paper presented at Proceedings of the 1998 IEEE/RSJ International Conference on Intelligent Robots and Systems. Part 1 (of 3), Victoria, Can, .

Learning-based automatic generation of collision avoidance algorithms for multiple autonomous mobile robots. / Fujita, Yukiyoshi; Fujita, Satoshi; Yamashita, Masafumi; Suzuki, Ichiro; Asama, Hajime.

1998. 1553-1558 Paper presented at Proceedings of the 1998 IEEE/RSJ International Conference on Intelligent Robots and Systems. Part 1 (of 3), Victoria, Can, .

Research output: Contribution to conferencePaper

Fujita, Y, Fujita, S, Yamashita, M, Suzuki, I & Asama, H 1998, 'Learning-based automatic generation of collision avoidance algorithms for multiple autonomous mobile robots' Paper presented at Proceedings of the 1998 IEEE/RSJ International Conference on Intelligent Robots and Systems. Part 1 (of 3), Victoria, Can, 10/13/98 - 10/17/98, pp. 1553-1558.
Fujita Y, Fujita S, Yamashita M, Suzuki I, Asama H. Learning-based automatic generation of collision avoidance algorithms for multiple autonomous mobile robots. 1998. Paper presented at Proceedings of the 1998 IEEE/RSJ International Conference on Intelligent Robots and Systems. Part 1 (of 3), Victoria, Can, .
Fujita, Yukiyoshi ; Fujita, Satoshi ; Yamashita, Masafumi ; Suzuki, Ichiro ; Asama, Hajime. / Learning-based automatic generation of collision avoidance algorithms for multiple autonomous mobile robots. Paper presented at Proceedings of the 1998 IEEE/RSJ International Conference on Intelligent Robots and Systems. Part 1 (of 3), Victoria, Can, .6 p.
@conference{9fd0174355b04261b33dcf3a528bdebf,
title = "Learning-based automatic generation of collision avoidance algorithms for multiple autonomous mobile robots",
abstract = "Using a model based on the omni-directional robots developed at the Institute of Physical and Chemical Research (RIKEN), we discuss the possibility of automatically generating a collision avoidance algorithm for autonomous mobile robots. To this end, we show that an effective collision avoidance algorithm for two robots can be generated by a very simple learning algorithm that simulates a naive human trial-and-error learning process, using only the robots' sensor outputs and a suitable reward function, where the exact form of the reward function is also learned autonomously by the robots. We also discuss how a robot can use its `experience' gained in a simple environment to adjust itself to a more complex environment, by automatically generating a collision avoidance algorithm for a three-robot situation utilizing a reduced state space resulting from the learning process for the case of two robots. The results of computer simulation and the experiments conducted at RIKEN using physical robots demonstrate the effectiveness of the collision avoidance algorithms generated and our learning-based approach.",
author = "Yukiyoshi Fujita and Satoshi Fujita and Masafumi Yamashita and Ichiro Suzuki and Hajime Asama",
year = "1998",
month = "12",
day = "1",
language = "English",
pages = "1553--1558",
note = "Proceedings of the 1998 IEEE/RSJ International Conference on Intelligent Robots and Systems. Part 1 (of 3) ; Conference date: 13-10-1998 Through 17-10-1998",

}

TY - CONF

T1 - Learning-based automatic generation of collision avoidance algorithms for multiple autonomous mobile robots

AU - Fujita, Yukiyoshi

AU - Fujita, Satoshi

AU - Yamashita, Masafumi

AU - Suzuki, Ichiro

AU - Asama, Hajime

PY - 1998/12/1

Y1 - 1998/12/1

N2 - Using a model based on the omni-directional robots developed at the Institute of Physical and Chemical Research (RIKEN), we discuss the possibility of automatically generating a collision avoidance algorithm for autonomous mobile robots. To this end, we show that an effective collision avoidance algorithm for two robots can be generated by a very simple learning algorithm that simulates a naive human trial-and-error learning process, using only the robots' sensor outputs and a suitable reward function, where the exact form of the reward function is also learned autonomously by the robots. We also discuss how a robot can use its `experience' gained in a simple environment to adjust itself to a more complex environment, by automatically generating a collision avoidance algorithm for a three-robot situation utilizing a reduced state space resulting from the learning process for the case of two robots. The results of computer simulation and the experiments conducted at RIKEN using physical robots demonstrate the effectiveness of the collision avoidance algorithms generated and our learning-based approach.

AB - Using a model based on the omni-directional robots developed at the Institute of Physical and Chemical Research (RIKEN), we discuss the possibility of automatically generating a collision avoidance algorithm for autonomous mobile robots. To this end, we show that an effective collision avoidance algorithm for two robots can be generated by a very simple learning algorithm that simulates a naive human trial-and-error learning process, using only the robots' sensor outputs and a suitable reward function, where the exact form of the reward function is also learned autonomously by the robots. We also discuss how a robot can use its `experience' gained in a simple environment to adjust itself to a more complex environment, by automatically generating a collision avoidance algorithm for a three-robot situation utilizing a reduced state space resulting from the learning process for the case of two robots. The results of computer simulation and the experiments conducted at RIKEN using physical robots demonstrate the effectiveness of the collision avoidance algorithms generated and our learning-based approach.

UR - http://www.scopus.com/inward/record.url?scp=0032313248&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0032313248&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0032313248

SP - 1553

EP - 1558

ER -