The Nomination for Competitive Co-Evolution of Sensory-Motor Systems by Gunnar Buason

I hereby nominate the following master's thesis for the SAIS Best AI Master's Thesis 2002 award:
"Competitive Co-Evolution of Sensory-Motor Systems" written by Gunnar Buason at the Department of Computer Science, University of Skövde (HS-IDA-MD-02-004; 116 pages; supervisor: Tom Ziemke)

Overview

This master's thesis investigates the use of competitive co-evolutionary algorithms for automatic design of robotic systems. More specifically, it presents twenty-one simulation experiments that integrate and extend previous work on competitive co-evolution of neural robot controllers in a predator-prey scenario with work on the 'co-evolution' of robot morphology and control systems. The focus is on a systematic investiga- tion of tradeoffs and interdependencies between morphological parameters and behavioral strategies through a series of predator-prey experiments in which increasingly many aspects are subject to self-organization through competitive co-evolution.

Motivations for nominating this thesis:

1. it is of high quality and well written;
2. it does a good job at integrating inspiration from evolutionary biology and cognitive science into cutting-edge AI research;
3. it is a nice example of basic AI research motivated by and feeding back to industrial applications of AI techniques (the work has been supported by and partly carried out at Ericsson Microwave Systems, Skövde);
4. last, but not least: the thesis advances the state of the art in its field; it has already resulted in one scientific publication ([1], a summary of the experiments (*)) to be presented at an evolutionary robotics workshop; a second paper has been submitted to the Genetic and Evolutionary Computation Conference (GECCO 2003), and a third one will be submitted to the European Conference on Artificial Life (ECAL 2003). (*) NB. That paper has also been submitted to the SAIS/SSLS meeting.

Reference and links:

[1] Buason & Ziemke (in press). Competitive Co-Evolution of Predator and Prey Sensory-Motor Systems. To appear in the proceedings of the Second European Workshop on Evolutionary Robotics (Colchester, UK, April 2003), Springer Verlag (Lecture Notes in Computer Science). [PDF, PS]

A 68-page appendix documenting in detail all experimental results is available as Technical Report HS-IDA-TR-02-004.

Gunnar Buason is continuing this work as a PhD student at the Department of Computer Science, University of Skövde, investigating the competitive co-adaptation of computer games and players.