MBARI’s Autonomous Systems Group, led by Kanna Rajan, has been working with scientists, engineers, and Marine Operations personnel to augment autonomous underwater vehicles (AUVs) with machine intelligence. Machine intelligence offers advancements in not only scientific research, but also AUV operations.
The Teleo-Reactive Executive (T-REX) adaptive control software system has been deployed routinely on AUV missions, successfully detecting dynamic features of interest, such as ocean fronts, and triggering water samplers within these hotspots. The system was also used to detect and map a sediment plume laden with agricultural runoff flowing into Monterey Bay. Yet another novel advance using this system enabled a scientist to retarget an AUV from shore.
After scientists received conductivity, temperature, and depth (CTD) data from the AUV via satellite, they directed the AUV to map and sample a particular front. T-REX received these goals and re-planned by turning the AUV around to lock onto the front’s hotspot. The long-term goal is to use such techniques to enable intelligent adaptive sampling targeted by a shore-side decision support system which integrates disparate data sources to aid scientists and operators(on ship or shore) where to direct one or more vehicles for further data collection and sampling.
Detailed description on Teleo-Reactive Executive (T-REX).
General information on autonomy and the T-REX architecture.