Fast homeostatic neural oscillators induce radical robustness in robot performance

  • Authors:

  • Ezequiel A. Di Paolo

  • Affiliations:

  • School of Cognitive and Computing Sciences, University of Sussex, Brighton, BN1 9QH, UK

  • Venue:
  • ICSAB Proceedings of the seventh international conference on simulation of adaptive behavior on From animals to animats
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

A network of relaxation oscillators is evolved to produce phototaxis in a simulated robot. Oscillations are faster than the timescale of performance, and are designed to maintain the same average activation value independently of sensory or synaptic input. Neural activation cannot correlate with any action-relevant sensory information, but must be continuously modulated by sensorimotor coupling. Radical sensor robustness is shown by inverting the position of the sensors and also by removing either of them in turn operations that do not alter the success of the strategy. Slowing down the timescale of oscillations results in less robustness.