Evolutionary online learning in multirobot systems

Abstract

Robots have the potential to replace manned machines and to carry out tasks in environments that are either remote or hazardous, such as space, deep sea, or underground. However, to create intelligent, reliable, mobile robots, capable of operating effectively in a wide variety of environments, the limited learning ability of robots needs to be addressed (Brooks & Matarić, 1993). Mobile robots are typically brittle: they are unable to adapt to changing environmental conditions (e.g., changes in terrain or lighting) or internal conditions (e.g., drift or permanent failure in sensors and/or actuators), and to learn new tasks as they execute.