Digital Twin Development of PEARL Autonomous Surface System Thermal Management

Hear about the development of PEARL, a solar-powered autonomous marine robotics platform that operates on the surface of small bodies of water and in the ocean. PEARL is solar-powered and can dock with autonomous underwater vehicles (AUVs) and provide real-time satellite connectivity through new satellite constellations such as Iridium^® NEXT and Starlink. As such, PEARL allows engineers to connect the oceans with space in real time. One of the key challenges, however, is thermal management. The platform experiences thermal extremes during both winter and summer. As a result of past thermal shutdown events, the PEARL team has developed a digital twin for thermal management that allows them to predict temperatures at six different locations within approximately 1% accuracy.

Discover the trade-off between the root mean square error (RMSE) of digital twin models and the complexity of the model itself in terms of the number of parameters that need to be modeled. You will also explore the broader lessons learned for predictive modeling of autonomous systems, which are applicable to other domains.