Distributed Robotic Intelligent Behaviors

Recent military efforts have illustrated the difficulty for conventional weapons to find and defeat "deep-targets" that are concealed, guarded, mobile, or even buried. Examples include underground hardened bunkers, cave/tunnel systems, and mobile SCUD launchers. Miniature robotics could be used to find, collect data, track, and even defeat such targets, but the robots need to be well coordinated, and possess an intelligence for creatively achieving their mission. A distributed robotics system architecture is proposed that allows a swarm of heterogeneous robots to be dropped into an area, and perform deep-target reconnaissance missions. Included in this approach are the collections of intelligent behaviors that control the individual robots, and the mechanisms for selection and composition of compound behaviors. The architecture takes into account the unique requirements and characteristics of the miniature robotic hardware, the communications, the power constraints, and types of locomotion and sensors that are anticipated to be present in the system. In Phase I, the distributed robotic software architecture was designed, and prototype behaviors are being implemented for analysis. During Phase 2, the system will be implemented and tests performed in simulation. Post Phase 2 plans include demonstration upon a collection of miniature robots already under development by our team.