NA23OAR0210327

In Phase I we developed a software capability for predicting the future positions of ocean drifters. We demonstrated the capability to predict drift for objects of varying shapes and train models based on historical observations. In Phase II we will leverage this software to perform high volume experimentation through which we will fully develop trained models using historical data for both passive drifters and uncrewed surface vehicles (USVs) with adjustable sails and rudders. We will perform a demonstration deployment of drifters and collect data to validate both our approach as well as the models we use as inputs. We will work with teams deploying and navigating USVs to perform deployment and route recommendations for a variety of mission types. Our work will result in predictive capabilities and tools that customers will use to more deliberately deploy, manage, and navigate their uncrewed assets. The predictive capabilities we develop will improve mission success for existing systems in scientific research, private sector, and military applications. We have documented the interest in our tools from manufacturers of unmanned systems and the end-users of those products. As these systems become increasingly prevalent, the demand for the tools we develop in this work will grow