Uncrewed Systems Technology 051 l Primoco One 150 l Power management l Ocius Bluebottle USV l Steel E-Motive robotaxi l UAVs insight l Xponential 2023 p Issue 51 Aug/Sept 2023 art 2 l Aant Farm TPR72 l Servos l Tampa Deep Sea Barracuda AUV

14 Platform one August/September 2023 | Uncrewed Systems Technology Saildrone is using Jetson modules with machine learning (ML) for its USVs to study weather, marine life and the ocean floor (writes Nick Flaherty). The nautical data collection technology has tracked hurricanes in the North Atlantic, discovered a 3200 ft underwater mountain in the Pacific Ocean and begun to help map the world’s ocean floor. The data streams are processed on Nvidia Jetson modules for AI at the edge and are being optimised in prototypes using Nvidia’s DeepStream software development kit (SDK) for intelligent video analytics. Saildrone is also using Nvidia’s JetPack SDK to run ML frameworks on the module for image-based vessel detection to aid navigation. The ML runs mostly locally on the Jetson module but can run on the cloud as well over a satellite connection, as bandwidth can be limited and costly to shuttle from its suite of high-resolution imaging sensors. The USVs have oceanographic sensors for measuring wind, temperature, salinity and dissolved carbon. Saildrone also enables research of ocean and lake floors using bathymetric sensors, including deep sonar mapping using single or multibeam sensors for going deeper or wider. Its perceptual sensor suite also includes radar and visual underwater acoustic sensors. DeepStream technology is used for image pre-processing and model inference on the vessel. Saildrone is seeking to make ocean intelligence collection cost-effective, offering data-gathering systems for science, fisheries, weather forecasting, ocean mapping and maritime security by providing more processing on the vessels. It has three different types of USV with a control centre service to monitor them and provide real-time data. All the USVs are monitored around the clock, and operators can change course remotely via the cloud if needed. Saildrone pilots set waypoints and optimise the routes using meteorological and oceanographic data from a vessel. “We’ve sailed into three major hurricanes, and right through the eye of Hurricane Sam, and the vehicles have come out the other side – they are pretty robust platforms,” said Blythe Towal, vice-president of software engineering at Saildrone. Running mostly on solar and wind power requires energy-efficient computing to handle so much data processing. “With solar power, being able to keep our computing load power efficiency lower than a typical computing platform running GPUs by implementing Nvidia Jetson is important for enabling us to do these kinds of missions,” said Towal. The University of Hawaii at Manoa is using three, 23 ft Saildrone Explorer USVs to study the impact of ocean acidification on climate change. A 6-month mission around the islands of Hawaii, Maui, Oahu and Kaui will help evaluate the ocean’s health, monitoring the impact of acidification on coral, oysters, clams, sea urchins and calcareous plankton. One of the Saildrone USVs recently completed a 370-day voyage monitoring carbon dioxide levels in the ocean, sailing from Rhode Island across the North Atlantic to Cabo Verde, down to the equator off the west coast of Africa, and back to Florida. Marine vessels ML helps ocean surveys Nvidia Jetson modules are helping to process Saildrone data about marine life and the ocean floor