A multi-agent system based on active vision and ultrasounds applied to fuzzy behavior based navigation

Abstract

In this paper, we present a multi-agent system that uses both visual and range sensors information to achieve a safe and efficient behavior-based navigation. The system uses a topological map of an indoor office-like environment and it is based on fuzzy behaviors, providing to the robot the ability to find doors in rooms. The system is formed by distributed agents that can establish communication among themselves. We use both reactive and deliberative agents and we have carried out a modular design of the system to facilitate its posterior expansion by adding new skills or new agents. Also, the use of a multi-agent system allows us to achieve a more robust performance of the robot. Regarding the behaviors, they have been designed using fuzzy rules to set the appropriate relationship between the input data and the control values to apply to the robot actuators. The system has been implemented in a Nomad 200 mobile robot and has been validated in numerous tests in a read office-like environment