Cadet Justin Robinson ’21 presented research on Kriging Variance and Human Factors to Steer an Unmanned Aircraft System Swarm to Survey a Radiation Field at the American Institute of Aeronautics and Astronautics (AIAA) SciTech Conference, where his work was published.
This study explores the benefits of an Unmanned Aerial Vehicle (UAV) swarm that utilizes Kriging to map out and collect data from fields of interest. Kriging Swarm is an algorithm that repeatedly seeks to find the next set of optimal data points for the swarm to survey to minimize uncertainty in a field of unknown and random samples. The implementation of Kriging in a swarming behavior to accomplish this was more effective than a baseline algorithm. This was determined by conducting radiation simulations in SCRIMMAGE. Results were analyzed using a Mean Sum Squared Error (MSSE) analysis of the uncertainty values at each predicted value that comes from Kriging across the entire field of radiation. An investigation was conducted to evaluate the effects of human factors on the algorithm. Considerations for operator supervisory control elements are presented to successfully implement this survey mechanism in real-world operations.
CDT Robinson first became interested in drones as a child while flying remote control helicopters. During his rising Yearling summer, he participated in an AIAD at Pratt&Millers where he studied how a swarm of drones could integrate with a vehicle. He believes that any advance in UAV technology is beneficial for both civilians and military personnel. He hopes to have opportunities in the future to work with UAV technologies again.