Cloud-based Ocean Digital Twins as a Pathway Toward Scalable Marine Robotic Operations

As AUV and ASV technologies are increasing their capabilities to support mission durations of weeks to months and a range of hundreds to thousands of kilometers, a fundamental question exists on how to plan missions, predict vehicles performance, and make real time decisions on vehicles routes to maximize battery-powered vehicle endurance while minimizing the risk of vehicle loss. In this work we merged ocean digital twins with AUV dynamics and battery models in a scalable cloud framework that allows to run, during a mission of the physical systems, large Monte Carlo scenarios and autonomously communicate via satcom to the AUV the ideal route. The system was demonstrated during a combined mission of an ASV Waveglider and AUV Remus600, successfully identifying and transmitting to the AUV the ideal path that minimizes time-to-target and battery usage in the challenging ocean currents environment of the New England Shelf near the Woods Hole area