Abstract
Concept learning in robotics is an extremely challenging problem: sensory data is often high-dimensional, and noisy due to specularities and other irregularities. In this paper, we investigate two general strategies to speed up learning, based on spatial decomposition of the sensory representation, and simultaneous learning of multiple classes using a shared structure. We study two concept learning scenarios: a hallway navigation problem, where the robot has to induce features such as “opening” or “wall”. The second task is recycling, where the robot has to learn to recognize objects, such as a “trash can”. We use a common underlying function approximator in both studies in the form of a feedforward neural network, with several hundred input units and multiple output units. Despite the high degree of freedom afforded by such an approximator, we show the two strategies provide sufficient bias to achieve rapid learning. We provide detailed experimental studies on an actual mobile robot called PAVLOV to illustrate the effectiveness of this approach.
Article PDF
Similar content being viewed by others
Avoid common mistakes on your manuscript.
References
Borenstein, J. (1996). Navigating Mobile Robots. A.K. Peters.
Brooks, R. (1986). A robust layered control system for a mobile robot. IEEE Journal of Robotics and Automation, 2(1).
Caruana, R. (1993). Multitask learning: A knowledge-based source of inductive bias. In Proceedings of the Tenth International Conference on Machine Learning (pp. 41–48). Morgan Kaufmann.
Cassandra, T., Kaelbling, L. & Kurien, J. (1996). Acting under uncertainty: Discrete bayesian models for mobilerobot navigation. In IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 963–972.
Connell, J. & Mahadevan, S. (1993). Robot Learning. Kluwer Academic Publishers.
Dorigo, M. (1996). Introduction to the special issue on learning autonomous robots. In M. Dorigo, editor, IEEE Transactions on SMC, volume 26, pp 361–364.
Engleberger, J. (1989). Robotics in Service. MIT Press.
Fahlman, S. (1988). Faster-learning variations on back-propagation: An empirical study. In Proceedings of 1988 Connectionist Models Summer School. Morgan Kaufmann.
Franklin, J., Mitchell, T. & Thrun, S. editors. (1996). Robot Learning. Number 2–3. Kluwer Academic Press. Special Issue of Machine Learning, 23.
Jain. A. (1989). Fundamentals of Digital Image Processing. Prentice Hall.
Khaleeli, N. (1997). A robust robot navigation architecture using partially observable semi-markov decision processes. Master's thesis, University of South Florida, Tampa, FL.
Klingspor, V., Morik, K. & Rieger, A. (1996). Learning concepts from sensor data. In J. Franklin, T. Mitchell, and S. Thrun, editors, Recent Advances in Robot Learning. Kluwer Academic Press. Special Issue of Machine Learning, 23.
Koenig, S. & Simmons, R. (1997). Xavier: A robot navigation architecture based on partially observable markov decision process models. In D. Kortenkamp, P. Bonasso, and Murphy. R., editors, AI-based Mobile Robots: Case-studies of Successful Robot Systems. MIT Press.
Kosaka, A. & Kak, A. (1992). Fast vision-guided mobile robot navigation using model-based reasoning and prediction of uncertainties. In IEEE/RSJ International Conference on Intelligent Robots and Systems, pages 2177–2186.
Mahadevan, S. (1994). Proceedings of the workshop on robot learning. Held in conjuction with the 11th International Conference on Machine Learning (IMLC '94). Online proceedings available at http://www.cs.cmu.edu/afs/cs.cmu.edu/user/mitchell/ftp/robot-learning.html.
Mahadevan, S. (1996). Machine learning for robots: A comparison of different paradigms. In Proceedings of the IROS '96 Workshop on Towards Real Autonomy.
Mahadevan, S. & Connell, J. (1992). Automatic programming of behavior-based robots using reinforcement learning. Artificial Intelligence, 55, 311–365.
McClelland, J. & Rumelhart, D. editors. (1986). Parallel Distributed Processing: Explorations in the Microstructure of Cognition. Bradford Books.
Mitchell, T. (1997). Machine Learning. McGraw Hill.
Nourbakhsh, I., Powers, R. & Birchfield, S. (1995). Dervish: An office-navigating robot. AI Magazine, 16(2), 53–60.
Pomerleau, D. (1990). Neural network based autonomous navigation. In C. Thorpe, editor, Vision and Navigation: The CMU Navlab. Kluwer Academic Publishers.
Quinlan, J.R. (1986). Induction of decision trees. Machine Learning, 1(1), 81–106.
Salganicoff, M., Ungar, L. & Bajcsy, R. (1996). Active learning for vision-based grasping. In J. Franklin, T. Mitchell, and S. Thrun, editors, Recent Advances in Robot Learning. Kluwer Academic Press. Special Issue of Machine Learning Journal, Vol 23.
Shavlik, J. & Dietterich, T. (1990). Inductive learning from preclassified training examples. In J. Shavlik and T. Dietterich, editors, Readings in Machine Learning. Morgan Kaufmann.
Sutton, R. & Barto, A. (1998). Reinforcement Learning: An Introduction. MIT Press.
Tan, M. (1993). Cost-sensitive learning of classification knowledge and its application to robotics. Machine Learning, 13(1), 7–33.
Thrun, S. & Mitchell, T. (1994). Learning one more thing. Technical Report CMU-CS–94–184, Carnegie-Mellon University.
Rights and permissions
About this article
Cite this article
Mahadevan, S., Theocharous, G. & Khaleeli, N. Rapid Concept Learning for Mobile Robots. Machine Learning 31, 7–27 (1998). https://doi.org/10.1023/A:1007432422702
Issue Date:
DOI: https://doi.org/10.1023/A:1007432422702