Advanced Replanning and Execution Tools for Heterogeneous Unmanned Systems (ARETHUS)

Advanced automation capabilities and human-computer interfaces are needed to enhance operator situation awareness and reduce the cognitive work of supervising unmanned vehicle teams in complex airspaces. To address this need, we will design, develop, and evaluate Advanced Replanning and Execution Tools for Heterogeneous Unmanned Systems (ARETHUS). Three core attributes characterize our approach. First, we will expand prior analyses of unmanned vehicle operations to address the operator’s need for observability and directability of automated planning behaviors when supervising multiple vehicles. Second, we will leverage this analysis to drive the design of: (1) advanced planning algorithms that incorporate both hard domain constraints and soft operator preferences during mission planning and replanning, and (2) ecological mixed-initiative mission displays that enhance operator awareness of the relative safety and efficacy of automated plans, while intuitively conveying affordances for directing the planning process. Third, we will extend existing COTS, GOTS, and in-house planning algorithms, display tools, and simulation environments to rapidly develop working prototypes of ARETHUS component algorithms and mission displays. These prototypes will enable us to demonstrate and evaluate different approaches and therefore truly explore the full range of challenges in managing unmanned vehicles in complex airspaces.