Onboard adaptive control of AUVs using automated planning and execution

In this paper we describe an integrated goal-oriented control architecture for onboard decision-making for AUVs. Onboard planning and execution is augmented by state estimation of perceived features of interest in the coastal ocean, to drive platform adaptation. The parti-tioned architecture is a collection of coordinated control loops, with a recurring sense, plan, act cycle and which allows for plan failures to be localized within a control loop and ensures a divide-and-conquer approach to prob-lem solving in dynamic environments.