CiberRato is a competition about autonomous robotic control purely focused on the software and algorithmic details. The NEATsqueak project aims to develop an approach to this challenge based on artificial neural networks.
We're developing a series of studies on the use of NEAT, a neuroevolution technique, to optimize several modules of a CiberRato agent.
We've published an article on the 14th Portuguese Conference on Artificial Intelligence, regarding the optimization of a path following module.
Who are the people involved?
When did this project start?
We started planning and developing this in October, 2008.