Coordinator (PI): Tomás Arredondo
Contributors: Tomás Arredondo, Wolfgang Freund, César Muñoz, Nicolás Navarro, Fernando Quirós
Start Date: 1st July 2005
End Date: 1st December 2009
This project study and develop behavior based mobile robots. Initially, the project focused on developing new algorithm and methods for mobile robot applications with reduced sensory capabilities or with low cost sensors and actuators. The principal methods that we use include genetic algorithms, genetic programming and fuzzy logic.
Since 2007, the project study state of art algorithms and architectures to determine capabilities and limitations in real world setups. In order to make this process more efficient we divide the process into two stages. In the first stage, we utilize mobile robot simulators. In the second stage, we adapt and develop new algorithms capable of completing tasks in real world scenarios.
The project was started by Prof. Tomás Arredondo, Nicolás Navarro, César Muñoz on the second half of 2005 after a seminar on artificial intelligence. For more information, please visit the project official web site here.
Example of different trajectories generated by our genetic algorithms and fuzzy motivations approach. Left, trajectories displayed in YAKS. Right, trayectories displayed using scilab.
Left, IRMA-II, our mobile robot platform. Right, Real-world results recorded by a ceilling camera