An artificial intelligence approach to the 1994 AUVS Unmanned Ground Robotics Competition

Abstract

This paper describes the artificial intelligence (AI) approach taken by the CSM team to the Second Annual AUVS International Unmanned Ground Robotics Competition held in Rochester, Michigan, May 20-22, 1994. Using this approach, the CSM entry Omnibot won first place for the six student team members. The AI approach relied on a reactive architecture to decompose the robot's functions into a set of reflexive behaviors, and the principle of action-oriented perception to guide the design of each individual behavior and its relationship to the others. The major contribution of this paper is a case study of a successful reactive behavioral design for an autonomous mobile robot.