Automatic detection of visible events in fusion reactors

Abstract

Renewable energies are the future, and the study of those is already in research and development. Fusion reactors can give way to a new world inside the nuclear energy, very different to the current one. In this project, we aim at automatically detecting and classifying visible events by applying deep learning. Using labeled data from previous projects, we can contribute to detect anomalies in videos of the reactor core. This will help us to better understand the reactor behaviour. Furthermore, we will provide a new way to readily detect anomalies that may damage the structure of the reactor.