Extensible collaborative autonomy using GAMS

Abstract

The future applications for multi-agent systems abound given the ever decreasing cost associated with the technology. However, making use of such multi-agent systems, including multirobot systems, is still a challenge. Most historical robotic control paradigms depended on preplanned movements or having one or more humans in direct control of each robot. These approaches are inefficient in managing available resources and do not adapt well to changes in the environment or missions. The Group Autonomy for Mobile Systems (GAMS) middleware seeks to address the software development problem in distributed collaborative autonomous systems. The GAMS project facilitates multi-agent collaboration through a focus on extensible platforms and algorithms, object-oriented interfaces to shared knowledge in the multi-agent system, and seamless, non-blocking communication between peers. Available mechanisms for incorporating new platforms and algorithms into the GAMS system are presented.