Autonomous Radiation Mapping

Recent advances in nuclear sensors, computer vision, artificial intelligence, machine learning, and unmanned aerial/ground vehicles and unmanned aerial/ground vehicles could benefit autonomous sensors capable of performing in various mission spaces, such as cargo inspection, wide-area search, facility assessment. This requires the systems to have real-time data fusion capabilities that merge signals from the on-board radiation sensor suite with contextual sensor signals to enable visualization, scene mapping, and geolocation. RMD and DRAPER team proposes to develop an autonomous radiation detection, identification and mapping system for missions requiring operation in close cluttered spaces related to cargo inspection and facility assessment. RMD will leverage its own expertise in radiation sensors in form of E2Mx family, and Draper will leverage its expertise in remote operations with autonomous navigation based on platform-independent algorithms for visual-inertial odometry, tracking, path planning, obstacle avoidance, and precision landing. In Phase I, a preliminary prototype will be assembled utilizing RMDĺs E2MV and a small drone like from DRAPER. A single system will be used for autonomous navigation in cluttered conditions around and between containers to provide radiation, scene, and geolocation mapping. The capability to operate in low light conditions and without GPS will be evaluated. The ability to provide a specific radionuclide radiation field map will be demonstrated. A fully autonomous system capable of localizing and mapping of radiation fields, especially in a swarm configuration, will numerous applications beyond non-proliferation or safeguards. Survey of areas where the nuclear radiation is expected or suspected will be the main application. Such areas include nuclear power plants, waste storage, contaminated areas, mining operations, etc.