Self-Adaptive Robotic Architectures

Student: John C. Georgas, UC Irvine/ISR

Advisor: Richard N. Taylor, UC Irvine/ISR

Abstract:

Robotic systems are natural candidates for the application of self-adaptive software principles; the domain of this class of systems demand that robotic platforms be easily evolved and adapted to perform new tasks as needs and environmental conditions change. A major challenge, however, is that the majority of current approaches to robotic architectures suffer from a marked lack of support for both the qualities and characteristics which enable and foster self-adaptation as well as specific technologies for creating self-adaptive robotic systems. This poster and demo will present the RAS architectural style - a style tailored to promote qualities necessary for self-adaptation - combined with PBAAM, which is a policy-based approach to developing self-adaptive systems, in an effort to begin addressing the challenges of the domain and progress toward self-adaptive robotic systems.

Bio:

John Georgas is a Ph.D. candidate at the University of California, Irvine under the supervision of Professor Richard N. Taylor. His research interests include self-adaptive software, software architecture and related tools and representations, architectural styles, systems modeling, and robotic architectures. He received his B.S. in computer science from California State Polytechnic University, Pomona and his M.S. in computer science from UCI.