Automated Mission Scheduling by Distributed Constraint Optimization for Collaborative and Shared Control of Unmanned Vehicle Systems

Perceptronics Solutions, Inc., with the USC Computer Science Department Teamcore Research Group and Charles River Analytics, and supported by Professor Mary (Missy) Cummings of MIT presents this proposal to develop an Automated Mission Scheduler (AMS). The AMS will automatically create a mission plan for heterogeneous, multiple unmanned vehicles performing joint actions under the control of collaborating operators. The objective is to support dynamic mission scheduling in response to high-level mission objectives, priorities, constraints. Accordingly, the plans will include allocation of required mission tasks, including specific assignments to vehicles as well as sequencing of individual vehicle and mission tasks to achieve the collective objectives. Output of AMS will be used as a decision aid to the collaborating operators as well as for direct control as a component of evolving mixed initiative interface modules concepts. Our team has already made significant progress toward achieving the project goals, and our proposed system is based on innovative technology products contributed by our team members: USC’s ground-breaking work to apply Distributed Constraint Optimization to optimally allocate multi-vehicle tasks in collaborative control of UVs, Charles River’s advanced Human/UV interface technology, and Perceptronics Solutions’ Tactical Group Decision Analysis System and Mixed Initiative Human-Robots Team Performance Assessment System.