The goal of Adaptive Robotics is to design and operate various platforms of mobile robotic systems that are able to navigate autonomously in open and changing environments and to cope with interruptions in communication links, component failures, and changes in workload, resources, and mission requirements. Systems are realized as wheel-driven robots and walking machines, as all-terrain vehicles or even as swarms of interactive robots.

Control of adaptive robots is provided by classical AI methods as well as by fuzzy logic techniques and neural networks. Adaptivity is generated by appropriate learning methods like reinforcement learning or evolutionary processes.

 

Contact:
Uwe Beyer
uwe.beyer(at)iais.fraunhofer.de
Phone +49 2241 14 3020