Adaptive learning of a robot arm

  • Mukesh J. Patel
  • Marco Dorigo
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 865)


Alecsys, an implementation of a learning classifier system (LCS) on a net of transputers was utilised to train a robot arm to solve a light approaching task. This task, as well as more complicated ones, has already been learnt by Alecsys implemented on AutonoMouse, a small autonomous robot. The main difference between the present and previous applications are, one, the robot arm has asymmetric constraints on its effectors, and two, given its higher number of internal degrees of freedom and its non anthropomorphic shape, it was not obvious, as it was with the AutonoMouse, where to place the visual sensors and what sort of proprioceptive (the angular position of the arm joints) information to provide to support learning. We report results of a number of exploratory simulations of the robot arm's relative success in learning to perform the light approaching task with a number of combinations of visual and proprioceptive sensors. On the bases of results of such trials it was possible to derive a near optimum combination of sensors which is now being implemented on a real robot arm (an IBM 7547 with a SCARA geometry). Finally, the implications these findings, particularly with reference to LCS based evolutionary approach to learning, are discussed.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Booker L., D.E. Goldberg and J.H. Holland (1989). Classifier systems and genetic algorithms, Artificial Intelligence, 40, 1–3, 235–282.Google Scholar
  2. Colombetti M. and M. Dorigo (1992). Learning to control an autonomous robot by distributed genetic algorithms. Proceedings of From Animals To Animats, 2nd International Conference on Simulation of Adaptive Behaviour (SAB92), Honolulu, HI, MIT Press, 305–312.Google Scholar
  3. Davidor Y. (1991). Genetic Algorithms and Robotics: A heuristic strategy for optimization. World Scientific: Singapore.Google Scholar
  4. Dorigo M. (1992). Using transputer to increase speed and flexibility of genetics-based machine learning systems. North Holland, Microprocessing and Microprogramming, Euromicro Journal, 34, 147–152.Google Scholar
  5. Dorigo M. (1993). Genetic and Non-Genetic Operators in ALECSYS. Evolutionary Computation Journal, 1, 2, 149–162.Google Scholar
  6. Dorigo M. (1994). ALECSYS and the AutonoMouse: Learning to Control a Real Robot by Distributed Classifier Systems. Machine Learning, forthcoming.Google Scholar
  7. Dorigo M. and M. Colombetti (1994). Robot Shaping: Developing Autonomous Agents through Learning. Artificial Intelligence, in press. Also available as Tech. Report No.92-040, International Computer Science Institute, Berkeley, CA.Google Scholar
  8. Dorigo M. and U. Schnepf (1993). Genetics-based Machine Learning and Behaviour Based Robotics: A New Synthesis. IEEE Transactions on Systems, Man, and Cybernetics, 23, 1, 141–154.Google Scholar
  9. Fogel D. B. (1994). Evolutionary Computation: Toward a New Philosophy of Machine Intelligence. IEEE Press, forthcoming.Google Scholar
  10. Holland J.H. (1975). Adaptation in natural and artificial systems. The University of Michigan Press, Ann Arbor, Michigan.Google Scholar
  11. Holland J.H. (1980). Adaptive algorithms for discovering and using general patterns in growing knowledge bases. International Journal of Policy Analysis and Information Systems, 4, 2, 217–240.Google Scholar
  12. Meyer J-A. and S.W. Wilson (1991). From Animals to Animats, Proceedings of the First International Conference on Simulation of Adaptive Behaviour, Bradford Books, MIT Press.Google Scholar
  13. Meyer J-A., H.L. Roitblat and S.W. Wilson (1993). From Animals to Animats 2, Proceedings of the Second International Conference on Simulation of Adaptive Behaviour, Bradford Books, MIT Press.Google Scholar
  14. Patel M.J. (1994a). Concept formation: A complex adaptive approach. Theoria 20 Google Scholar
  15. Patel M.J. (1994b). Situation assessment (S-A) and Adaptive Learning — Theoretical and Experimental Issues. Proceedings of The Second International Roundtable on Abstract Intelligent Intelligence. February, Rome, ENEA Headquarters.Google Scholar
  16. Varela F.J. and P. Bourgine (1992). Toward a Practice of Autonomous Systems, Proceedings of the First European Conference on Artificial Life, Bradford Books, MIT Press.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1994

Authors and Affiliations

  • Mukesh J. Patel
    • 1
  • Marco Dorigo
    • 1
    • 2
  1. 1.Progetto di Intelligenza Artificiale e Robotica, Dipartimento di Elettronica e InformazionePolitecnico di MilanoMilanoItaly
  2. 2.IRIDIAUniversité Libre de BruxellesBruxellesBelgium

Personalised recommendations