An Asynchronous, Robust, and Distributed Multisensor Fusion System for Mobile Robots

Abstract

In this paper, a multisensor fusion system that is used for calculating the position and orientation of an autonomous mobile robot is presented. The developed fusion system is distributed, robust, and asynchronous. It is distributed to permit the parallel function of all the sensors. It is robust because, being distributed, the system has been designed to keep working properly in spite of the failure, removal, or change of any sensor. It is asynchronous to take advantage of the different features and rates of each sensor. The implementation of the system is based on the distributed Kalman filter developed by Durrant-Whyte and Rao. In that distributed filter, all the sensors work in parallel to obtain their own estimate based on their own observations and on the observations coming from other sensors. Changes have been made to simplify and speed up the computation of the external validation equations and to allow the use of any sensor model. The equations have also been adapted to deal with asynchronously operating sensors and with the existence of communication delays. The fusion system is used to estimate the position and orientation of a mobile robot. The performance of the fusion system is shown both under simulation and in a real test with a mobile robot.