This research was supported by EU-Cognition for a period of five months after finishing my final year project. During the last two months we researched the possibility of improving the previous system that resulted from my final year project. The major limitation of my final year project was the lack of generalisation of the evolved neural network controller. In other words, the rover was not reliable enough in all sorts of different terrains. In addition, the number of sensors that were previously used was rather high and therefore our intention during this period was to achieve minimal number of these infrared sensors while keeping the same level of sensory feedback from the environment. We have changed the environment, the way the rover's behaviour was evaluated and also minimised the number of sensors by double thanks to the reuse of ground sensors for both the detection of obstacles and also holes.
Abstract from the publication that I presented at European Space Agency during the Advanced Space Technologies for Robotics and Automation (ASTRA 2008) conference
The paper presents an evolutionary robotics model of the Rover Mars robot. This work has the objective to investigate the possibility of using an alternative sensor system, based on infrared sensors, for future rovers capable of performing autonomous tasks in challenging planetary terrain environments. The simulation model of the robot and of Mars terrain is based on a physics engine. The robot control system consists of an artificial neural network trained using evolutionary computation techniques. An adaptive threshold on the infrared sensors has been evolved together with the neural control system to allow the robot to adapt itself to many different environmental conditions. The properties of the behaviour obtained after the evolutionary process has been tested by measuring the performance of the rover under various terrain conditions. Simulations results show that the robot, at the end of the evolutionary process, is able to avoid rocks, holes and steep slopes based purely on the information provided by the infrared sensors.
To download full paper click here