SBIR-STTR Award

Physical Sciences Inc. is proposing to develop the Network Fusion Radiation Analysis Tool (NFRAT) to address the need for an enhanced detection, localization and identification capability in support of sensitive site exploitation, perimeter monitoring and contamination avoidance operations. The NFRAT software will utilize probabilistic representations of the radiological environment and statistical models of radiation detector response to fuse measurements from multiple radiation detectors reported over a tactical network. The NFRAT algorithms will combine the continuously reported data from non-spectroscopic sensors, such as dosimeters, with less frequent, but more specific data from spectroscopic detectors to update a probability density function (PDF) of source parameters. The knowledge accumulated within the PDF will be continuously and automatically analyzed to generate actionable estimates of source parameters. The NFRAT fusion approach will be designed to reduce the time needed to characterize a source by more than 30% compared to non-networked approaches. NFRAT will generate source isotopic identification, activity, location and motion parameters. NFRAT will also detect the presence of distributed sources and estimate a spatial map of contamination levels and type. NFRAT will be integrated with situational awareness tools, such as Team Awareness Kit (TAK), to provide visual feedback on the search status.

Physical Sciences Inc. proposes to develop and mature the Network Fusion Radiation Analysis Tool (NFRAT) to address the need for an enhanced radiation hazard mapping, localization and identification capability in support of contamination avoidance operations and sensitive site exploitation. The NFRAT software will utilize dynamic representations of the radiological environment and statistical models of radiation detector response to fuse measurements from multiple radiation detectors. The NFRAT algorithms will combine continuously reported data from non-spectroscopic sensors, such as dosimeters, with less frequent, but more specific, data from spectroscopic detectors to update an algebraic system of constraints on source parameters. The knowledge accumulated within the system will be automatically analyzed to generate actionable maps of radiation fields due to distributed and compact sources as well as detect, localize and identify point sources. The NFRAT fusion approach will be designed to reduce the time needed to characterize the radiation field by more than 30% compared to non-networked approaches while providing fused dose rate estimates with < 25% relative error. NFRAT will also compute uncertainty of the estimated radiation field parameters. NFRAT will be integrated with situational awareness tools, such as Tactical Assault Kit (TAK), to provide visual feedback on the search status.