A Cognitive Architecture for Naval Mine Countermeasures (MCM)

Mine Countermeasures (MCM) missions of tomorrow will increasingly exploit autonomous vehicles such as the remote multi-mission vehicle to reduce risk to personnel and equipment. However, existing approaches are limited for continuous sensing and planning needed to deal with the dynamic and uncertain nature of the MCM missions. To address this gap we will consider alternative approaches that merge symbolic and probabilistic reasoning, a complementary approach to existing mathematical optimization, to handle planning and replanning under uncertainty. We will develop and investigate a system called CAMPER (A Cognitive Architecture for MCM planning, execution, and replanning). CAMPER will respond effectively to evolving MCM environments by making and executing plans that coordinate multiple autonomous vehicles to achieve goals. This will form the basis of a promising operator friendly system that addresses the problem of decision support and automation in high-risk, high-uncertainty MCM missions. As part of CAMPER, we will investigate novel methods for deliberative planning under uncertainty, environment monitoring using distributed sensors, and explanation as a means of understanding the environment and engendering trust. We will demonstrate the feasibility of CAMPER by evaluating it on simulated missions.