Gaussian Processes for radiation dose prediction in nuclear power plant reactors

In nuclear power plants, there are high-exposure jobs, like refuelling and maintenance, that require getting close to the reactor between operation cycles. Therefore, reducing radiation dose during these periods is of paramount importance regarding safety regulations. While some manipulable variables, levels certain corrosion products, can influence final level dose, no way determine it in a principled way. this work, we propose use Machine Learning predict at cycle end based on information available operation. particular, Gaussian Process model relation cobalt radioisotopes (a kind product) levels. Processes acknowledge uncertainty their predictions, desirable property considering high-risk nature present application. We report experiments real data gathered from five different plants Spain. Results show models be used estimate future values data-driven Moreover, tools currently development for application plants.