I hereby would like to nominate the following master's thesis for the SAIS Best AI Master's Thesis 2001 award:
"Exploring Internal Simulation of Perception in a Mobile Robot" written by Dan-Anders Jirenhed, Cognitive Science Program, IDA, Linköping University (LIU-KOGVET-D-0047-SE; supervisor: Tom Ziemke, Högskolan Skövde)
This master's thesis uses AI (evolutionary algorithms and artificial neural networks for learning in and control of a mobile robot) to model something that plays a central role for natural, in particular human, intelligence, namely the capacity to make use of an 'inner world'. Traditionally, AI and cognitive science have explained this phenomenon with reference to internal models and representations.
Dan-Anders' thesis, on the other hand, integrates neuro- scientific theories, philosophical arguments and adaptive robotics research to develop and test experimentally an alternative, non-representational explanation based on agents' capacity to internally simulate their own interaction with the world.
Motivations for nominating this thesis:
[1] Jirenhed, Hesslow & Ziemke (2001). Exploring Internal Simulation of Perception in Mobile Robots. In: Arras, Baerveldt, Balkenius, Burgard, Siegwart (eds.) 2001 Fourth European Workshop on Advanced Mobile Robotics - Proceedings (pp. 107-113). Lund University Cognitive Studies, vol. 86. Lund, Sweden.