Abstract
If autonomous robots are to act “intelligently”, they need to be equipped with powerful planning and control capabilities. Planning is an off-line activity that entails reasoning symbolically upon an explicit representation of the environment and the task to be performed. Control is an on-line activity, where an actual motion is made to adjust as closely as possible to a commanded motion through the use of sensory feedback. It is claimed in this paper that planning is amenable to symbolic central processing, while control is better realized through subsymbolic distributed computing. This distinction carries on to the learning domain, so that the acquisition of planning knowledge would be done symbolically at a central level, while the learning of reflexes and skills would be carried out at a neural peripheral level. After discussing this claim and providing support for it, we look at its implications. Thus, two traditionally disjoint research fields, namely Geometric Motion Planning and Neural Robot Control, become related in this context. Some of our works in these two fields are described under this unifying perspective, our final aim being to devise knowledge and data representations that permit interfacing geometric planners with neural controllers.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Arbib M.A. (1981): Perceptual structures and distributed motor control. In Handbook of Physiology - The Nervous System II. Motor Control, edited by V.B. Brooks, pp. 1449–1480, American Physiological Society, Bethesda, MD.
Avnaim F., Boissonnat J.D. and Faveijon B. (1988): A practical motion planning algorithm for polygonal objects amidst polygonal obstacles. IEEE Conf. on Robotics and Automation, Philadelphia, pp. 1656–1661.
Barhen J., Gulati S. and Zak M. (1989): Neural learning of constrained nonlinear transformations. IEEE Computer, June, pp. 67–76.
Barto A.G. (1985): Learning by statistical cooperation of self-interested neuron-like computing elements. Human Neurobiology 4, pp. 229–256.
Barto A.G. and Sutton R.S. (1981): Landmark learning: an illustration of associative search. Biological Cybernetics 42, pp. 1 - 8.
Brooks R.A. (1983): Solving the find-path problem by good representations of free space. IEEE Trans, on Systems, Man, and Cybernetics 13, pp. 190–197.
Brooks R.A. and Lozano-Perez T. (1985): A subdivision algorithm in configuration space for findpath with rotation. IEEE Transactions on Systems, Man, and Cybernetics 15, pp. 224–233.
Canny J.F. (1988): The Complexity of Robot Motion Planning. MIT Press.
Ewert J.-P. and Arbib M.A. (eds.) (1989): Visuomotor Coordination: Amphibians, Comparisons, Models, and Robots. Plenum Press, New-York London.
Goldberg K. and Pearlmutter B. (1988): Using a neural network to learn the dynamics of the CMU direct-drive arm n. Technical Report CMU-CS-88-160, Computer Science Department, Carnegie-Mellon University.
Gouzenes L. (1983): Generation of collision-free trajectories for mobile and manipulator robots. Technical report, LA AS du CNRS, Toulouse.
Graf D.H. and LaLonde W.R. (1988): A neural controller for collision-free movement of general robot manipulators. Proc. 2nd IEEE Intl. Conf. on Neural Networks, Vol. I, pp. 77–84.
Grossberg S. and Kuperstein M. (1986): Neural Dynamics of Adaptive Sensory-motor Control: Ballistic Eye Movements. Elsevier, Amsterdam.
Guez A. and Selinsky J. (1988): A trainable neuromorphic controller. Journal of Robotic Systems 5 (4), pp. 363–388.
Hecht-Nielsen R. (1990): Neurocomputing; Addison-Wesley.
Hinton G. (1989): Connectionist learning procedures. Artificial Intelligence 40, pp. 185–234
Ilari J. and Torras C. (1990): 2D path planning: a configuration space heuristic approach. The Intl. Journal of Robotics Research 9, pp. 75–91.
Kawato M., Uno Y., Isobe M. and Suzuki R. (1987): A hierarchical model of voluntary movement and its application to robotics. Proc. IEEE 1st Intl. Conf. on Neural Networks, San Diego.
Khatib O. (1986): Real time obstacle avoidance for manipulators and mobile robots. The Intl. Journal of Robotics Research 5, pp. 90–98.
Kohonen T. (1988): Self-Organization and Associative Memory ( second edition ). Springer-Verlag, Berlin Heidelberg New-York Tokyo.
Kung S-Y. and Hwang J-N. (1989): Neural network architectures for robotic applications. IEEE Trans, on Robotics and Automation 5 (5), pp. 641–657.
Kuperstein M. (1987): Adaptive visual-motor coordination in multijoint robots using a parallel architecture. Proc. of the IEEE Intl. Conf. on Robotics and Automation, pp. 1595–1602.
Latombe J-C. (1991): Robot Motion Planning, Kluwer Academic Publishers, Boston.
LeCun Y. (1985): Une procedure d’aprentissage pour reseau au seuil assymetrique. Proc. of Cognitiva, pp. 599 - 604.
Lozano-Perez T. and Wesley M. (1979): An algorithm for planning collision-free paths among polyhedral obstacles. Communications of the ACM 22(10), October, pp. 560–570.
Mel B.W. (1989): MURPHY: a neurally-inspired connectionist approach to learning and performance in vision-based robot motion planning. Technical Report CCSR-89-17, Center for Complex Systems Research, University of Illinois at Urbana-Champaign.
Millan J. del R. and Torras C. (1992): A reinforcement connectionist approach to robot path finding in non-maze-like environments. Machine Learning, in press.
Miller W.T. (1989): Real-time application of neural networks for sensor-based control of robots with vision. IEEE Trans, on Systems, Man and Cybern. 19(4), pp. 825–831.
Miller W.T., Sutton R.S. and Werbos RJ. (1990): Neural Networks for Control, The MIT Press, Cambridge and London.
O’Dunlaing C. and Yap Ch.K. (1985): A “retraction” method for planning the motion of a disc. Journal of Algorithms 6, pp. 104–111.
Pao Y-H., Sobajic D.J. (1987): Artificial neural-net based intelligent robotics control. Proc. 6th SPIE Conf. on Intelligent Robots and Computer Vision, pp. 542–549.
Reeke G.N. and Edelman G.M. (1987): Selective neural networks and their implications for recognition automata. Intl. Journal of Supercomputing Applications 1, pp. 44–69.
Ritter H. and Schulten K. (1988): Extending Kohonen’s self-organizing mapping algorithm to learn balistic movements. In Neural Computers, edited by R. Eckmiller and C. von der Malsburg, ASI NATO Series F: Computer and Systems Sciences 41, Springer-Verlag, Berlin Heidelberg New-York Tokyo.
Rumelhart D.E., Hinton G.E. and Williams R.J. (1986): Learning representations by back-propagating errors. Letters to Nature 323, pp. 533–535
Schwartz J. T. and Sharir M. (1983): On the “piano movers” problem. II. General techniques for computing topological properties of real algebraic manifolds. Advances in Applied Mathematics 4, pp. 298–351.
Sutton R.S. (1990): Integrated architectures for learning, planning and reacting based on approximating dynamic programming. Proc. 7th Annual Conf. of the Cognitive Science Society.
Torras C. (1989): Relaxation and neural learning: points of convergence and divergence. Journal of Parallel and Distributed Computing 6, pp. 217–244.
Torras C. (1990): Motion planning and control: Symbolic and neural levels of computation. Proc. 3rd. COGNITIVA, Madrid, November, pp. 207–218.
Widrow B. and Hoff M.E. (1960): Adaptative switching capability and its relation to the mechanisms of association. Kybernetik 12, pp. 204–215.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1993 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Genís, C.T. (1993). Symbolic Planning versus Neural Control in Robots. In: Rudomin, P., Arbib, M.A., Cervantes-Pérez, F., Romo, R. (eds) Neuroscience: From Neural Networks to Artificial Intelligence. Research Notes in Neural Computing, vol 4. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-78102-5_30
Download citation
DOI: https://doi.org/10.1007/978-3-642-78102-5_30
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-56501-7
Online ISBN: 978-3-642-78102-5
eBook Packages: Springer Book Archive