Evolutionary Spiking Circuits for Vision-Based Robots

Dario Floreano
Autonomous Systems Laboratory
Institute of Systems Engineering
Swiss Federal Institute of Technology, Lausanne

This is the abstract of a talk prepared for the International interdisciplinary seminar on new robotics, evolution and embodied cognition (IISREEC).12th to 15th November 2002, Lisbon, Portugal

Abstract: I will give an overview of our recent work on evolution of spiking neural circuits for control of a range of autonomous mobile robots. All the evolutionary experiments are carried out on physical robots without human intervention. After discussing how to implement and interface these neurons with a physical robot, I will show that evolution finds relatively quickly functional spiking controllers capable of navigating in textured environments without hitting obstacles using very simple genetic encoding and fitness functions. Neuro-ethological analysis of the network activity provides insight into the functioning of evolved controllers and tell the relative importance of single neurons independently of their observed firing rate. A number of systematic lesion experiments indicate that evolved spiking controllers are very robust to synaptic strength decay that typically occurs in hardware implementations of spiking circuits. Capitalizing on these results, I will then describe current work on evolution of spiking circuits in wheeled micro-robots (1 cm), in vision-based flying robots, and in the "electronic tissue", a new breed of evolvable hardware.

Related Links


For more information contact Luis Rocha at rocha@lanl.gov
Last Modified: October 17, 2002