A reactive robot navigation system based on a fluid dynamics metaphor

Abstract

Mobile robots need a powerful and flexible navigation system in order to move around autonomously in an incompletely known and changing environment. In this paper we present a computational metaphor for path generation which is gleaned from fluid dynamics. It is powerful because it can find optimal paths in a maze of arbitrary complexity and it is flexible because it readily adapts to any change in the topology of the maze. Moreover, it can be proven that the path generator does not suffer from local minima, a defect which hampers some of the other metaphor based methods. We show that the fluid dynamics metaphor also provides ample possibilities to solve more complicated navigation problems including navigation through weighted regions. The paper discusses the natural parallelism of the metaphor and its implementation on the DAP computer, which is an SIMD machine.

This research has been sponsored by the Belgian Government with the contract ”Incentive Program For Fundamental Research In Artificial Intelligence ; Project : Self-organization in subsymbolic computation” and ”Geconcerteerde Actie : Artificiële intelligentie, Parallelle Architecturen en Interfaces”. Part of this research has also been funded by the Esprit Program with the contract P440: ”Message Passing”