Coordinating UAVs in Dynamic Environments by Network-Aware Mission Planning

Traditional AI planning has been used successfully in many domains, including logistics, scheduling and game playing. This paper examines how AI planning techniques can be extended to coordinate teams of unmanned aerial vehicles (UAVs) in dynamic environments. Specifically challenging are real-world environments where UAVs and other network-enabled devices must communicate to coordinate -- and communication actions are neither reliable nor free. Such network-centric environments are common in military, public safety and commercial applications, yet most planning research (even multi-agent planning) usually takes communications among distributed agents as a given. The emerging application challenge of unmanned systems makes this problem of central focus. This work examines the problem of planning, plan monitoring and coordination of the mission of multiple UAVs in a communication-constrained environment. The work introduces several abstractions that enable AI planners to reason about communication and networking knowledge, and provides the underlying network system the means for including mission data as part of network operations. This work has been empirically validated using a distributed network-centric software evaluation test bed and the results provide guidance to designers in how to understand and control intelligent systems that operate in these environments.