Directed sonar sensing for mobile robot navigation. (English) Zbl 0760.68008

The Kluwer International Series in Engineering and Computer Science. 175. Dordrecht etc.: Kluwer Academic Publishers. XIX, 183 p. (1992).
The problem of mobile robot navigation using sonar is investigated. Model-based navigation is viewed as a process of tracking naturally occurring environment features, and named “targets”. Targets that have been predicted from the environment map are tracked to provide an estimated vehicle position. Targets that are observed, but not predicted, represent unknown environment features or obstacles, and cause new tracks to be initiated, classified, and integrated into the map.
An algorithm for model-based localization that is based on an extended Kalman filter that utilizes matches between observed and predicted features from an a priori map to update a position is described. Problem of the bottom-up interpretation of unexpected observations to provide obstacle detection and map building capabilities are investigated. Experimental results are obtained for the restricted case of learning an unknown room from precisely known vehicle locations. Localization while learning and map maintenance in a dynamic environment and the concept of directed sensing are discussed.
Reviewer: Y.V.Krak (Kiev)


68-02 Research exposition (monographs, survey articles) pertaining to computer science
68U99 Computing methodologies and applications