Neural network applications in sensor fusion for an autonomous mobile robot
In this article, we propose a generic architecture for sensor data fusion and argue that the central issue in such an approach is the choice of a suitable representation of the robot's environment. We argue that for the navigation task a robot-centered discrete probabilistic representation (an occupancy grid) is a suitable choice. If such a representation is used, the two key problems are how to transform such representations upon robot motion and how to represent the sensor's error characteristics (the sensor model) in such a representation. For both these problems, solutions are suggested by the application of neural network theory, and it is argued that these neural networks are the best available alternatives.
keywordssensor fusion neural networks occupancy grids transformation of occupancy grids learning sensor models
Unable to display preview. Download preview PDF.
- 1.J.T. Schwartz, M. Sharir, “On the Piano Movers' Problem. I. The Case of a Two-Dimensional Rigid Polygonal Body Moving Amidst Polygonal Barriers”, Communications on Pure and Applied Mathematics, Vol. XXXVI, 193, pp. 345–398.Google Scholar
- 2.A. Elfes, “Using Occupancy Grids for Mobile Robot Perception and Navigation”, IEEE Computer, June 1989, pp. 46–57Google Scholar
- 3.J. del R. Millan, “Reinforcement Learning of Goal-Directed Obstacle-Avoiding Reaction Strategies in an Autonomous Mobile Robot”, Technical note, March 1993Google Scholar
- 4.D.J. Braunegg, “MARVEL: A System That Recognizes World Locations with Stereo Vision”, IEEE Trans, on Rob. and Aut., vol 9, no 3, June 1993, pp. 303–308.Google Scholar
- 5.Hugh F. Durrant-Whyte Integration, Coordination and Control of Multi-Sensor Robot Systems. Kluwer Academic Publishers, 1988Google Scholar
- 6.Y. Shirai, “Visual Sensor Fusion”, Proc. of 1994 Int. Conf on Multisensor Fusion and Integration for Intelligent Systems, Las Vegas, Nevada, October 1994. Tutorial.Google Scholar
- 7.G. Shafer, “A Mathematical Theory of Evidence”, Princeton University Press, Princeton 1976.Google Scholar
- 8.A. Sabater, “Set Membership Approach to the Propagation of Uncertain Geometric Information”, Proc. of 1991 IEEE Int. Conf. on Rob. and Aut., Sacramento, California, April 1991, pp 2718–2723Google Scholar
- 9.Kröse, B.J.A. and E. Dondorp,“A Sensor Simulation System for Mobile Robots”, in: T. Kanade, F.C.A. Groen and L.O. Hertzberger (ed.), Intelligent Autonomous Systems 2, December 1989.Google Scholar
- 11.J.W.M. van Dam, B.J.A. Kröse and F.C.A. Groen, “Transforming the Ego-centered Internal Representation of an Autonomous Robot with the Cascaded Neural Network”, Proc. of 1994 Int. Conf on Multisensor Fusion and Integration for Intelligent Systems, Las Vegas, Nevada, October 1994Google Scholar
- 12.Alberto Elfes, “Sonar-based real world Mapping and Navigation”, IEEE Journal of robotics and automation, VOL. RA-3, NO. 3, June 1987.Google Scholar
- 13.Tom Brotherton, “Hierarchical Neural Net Processing”, IEEE Int. Conf. on Multisensor Fusion and Integration for Intelligent Systems, Las Vegas, Nevada, October 1994. Tutorial.Google Scholar