The Artificial Evolution Of Control Systems

Recently there have been a number of proposals for the use of artificial evolution as a radically new approach to the development of robot control systems. This paper explains and justifies the evolutionary robotics methodology developed and used at Sussex University. Results are presented from research into the concurrent evolution of dynamical network controllers and visual sensor morphologies for an autonomous mobile robot. Successful experiments to evolve robot controllers directly in hardware, thus allowing advantage to be taken of intrinsic silicon physics, are also discussed. The implications of recent work in using high quality engineering simulations to evolve control systems for real robots will are covered. INTRODUCTION When designing a control system for a robot, there are at least three major problems: ffl It is not clear how a robot control system should be decomposed. ffl Interactions between separate sub-systems are not limited to directly visible connecting links, ..