Development of a Semi-Autonomous Aerial Radiation Detection System

This paper outlines the integration of the Broad-Area Search Bayesian Processor with a networked, unmanned, aerial radiation detection system. The advent of advanced machine learning coupled to a semi-autonomous search platform affords the opportunity to significantly increase the effectiveness of broad-area radiological search operations. Technological advances in the realm of radiation detection systems tend to focus on improving system parameters such as efficiency or resolution. Operationally, however, approaches to the broad-area search problem remain tethered to the standard gridded search pattern. Increasing the range at which systems detect, localize, and identify or characterize radioactive materials while coupling the detector data with prior knowledge of the search space offers the ability to employ advanced data processing not previously available to radiation detection platforms. In this paper, we focus on the development of a small, low-cost, unmanned, aerial radiation detection platform and its integration with our Broad-Area Search Bayesian Processor. While we demonstrate the utility of employing such systems based on a single-system, we envision the employment of multiple, networked sensors (both ground-based and airborne) to significantly decrease the time required to reduce the search space.