Artificial evolution of FPGA that controls a miniature mobile robot Khepera

摘要

We here demonstrate an artificial evolution of a hardware that controls a real autonomous mobile robot by the microbial genetic algorithm (GA). An add-on turret of the robot with a field programmable gate array (FPGA, XILINX XC6216) was used as the evolvable hardware. The mobile robot with the FPGA successfully evolved to acquire reasonably good navigation capabilities within an acceptable period of time. We exhibit that the microbial GA provide better chance to get higher score than the conventional GA.