A Cognitive Architecture for Replanning and Execution (CARPE)

Mine countermeasures (MCM) missions require complex coordination of personnel, equipment, and autonomous vehicles. The mission success hinges on timely and accurate decision-making for responding to dynamic situations in the area of operations that pose considerable risks to mission assets; both personnel and equipment. However, no decision support tools are presently available for mission personnel to exploit the most up to date situation information and maximize mission performance. To fill this capability gap, we will develop a decision support system called CARPE; a Cognitive Architecture for RePlanning and Execution, that shall support decision making for MCM missions with replanning, retasking, and rescheduling capabilities, and shall interoperate with the MEDAL-EA system. We will develop novel techniques for deviation detection, interactive deviation cause analysis, minimally disruptive rescheduling, causal and plan model learning. We will implement and integrate these technologies as reasoning services with MEDAL-EA to provide a seamless decision support experience to MCM mission personnel. We will deliver progressively mature versions of CARPE that will be evaluated by Subject Matter Experts using simulation assisted decision-making.