Towards a distributed, cognitive robotic architecture for autonomous heterogeneous robotic platforms
Contributing USMA Research Unit(s)
Electrical Engineering and Computer Science, Robotics Research Center
This paper describes an early phase towards a general, holistic, distributed robotic architecture capable of supporting different domain specific tasks autonomously or semi-autonomously on multiple platforms. Rather than focusing on individual components and algorithms, we take a systems level approach striving for a task-independent architecture. Our ultimate goal is to employ autonomous scouts that augment a military unit with additional reconnaissance and security assets. We envision one solider controlling a mix of three to five ground and aerial robots in contrast to one solider controlling a single robot as realized in currently deployed systems. Such a general robotic architecture may be useful in other domains to include law enforcement, search, and rescue, hospital logistics, and domestic use. To demonstrate interoperability this paper discusses the architectural design and current set of heterogeneous prototypes built with common hardware, software, and a distributed, Web-based operator control unit (OCU). We explore future extensions that incorporate a cognitive architecture and work towards the Joint Architecture for Unmanned Systems (JAUS) compliance.
Lathrop, Scott D. and Korpela, Christopher, "Towards a distributed, cognitive robotic architecture for autonomous heterogeneous robotic platforms" (2009). USMA Research Papers. 16.