The main objective of this project is to detect, characterize, and locate radioactive sources in urban environments using computational models based on machine learning and statistical techniques. The project explores multiple approaches such as signal processing methods, and neural networks. Unnatural radiation sources, such as Uranium or Plutonium, can present a risk to the population if they remain undetected by radiological search and response teams. Moreover, the computational model being developed must be capable of identifying the type of radiation source, classifying it as innocuous (i.e., isotopes used in medical and industrial settings) or harmful (nuclear weapons). The project is currently supported by the Pacific Northwest National Laboratory (PNNL), in collaboration with the Department of Mathematics at Embry-Riddle.