Enhancement of AUV autonomy using backseat driver control architecture

The Deepwater Horizon oil spill accident (April 2010) in the Gulf of Mexico released an unprecedented amount of crude oil, much of which was trapped around 1000m below the surface. The accident has attracted worldwide attention and promoted many autonomous underwater vehicles (AUV) projects. The principal objective of our project is to establish a Backseat Driver control architecture for an AUV to monitor undissolvable marine pollutants. Our scope of work focuses on validating a new method to manoeuvre and control an AUV by implementing the Missions Oriented Operating Suite (MOOS-IvP) Backseat Driver system that was newly integrated on the Memorial University of Newfoundland (MUN) Explorer AUV. A Ping360 scanning sonar, as the main in-situ sensor, was integrated with the vehicle. This enabled a capability to survey horizontal sectors up to 360° and ranges up-to 50m from the AUV in real-time to evaluate information of sonar reflections from objects surrounding the vehicle without a human-operator-in-the-loop. We validated the capability of this intelligent Backseat Driving control through three sets of field experiments that were conducted in October – November 2020 in Newfoundland, Canada. 