Paul Boniol, a post-doctoral researcher in computer science has won several awards for his work, which lead to the filing of 2 patents at both the national and European level. He explains his methods for analysing large volumes of data to create algorithms capable of predicting all kinds of anomalies.

Paul Boniol did his research at LIPADE, in collaboration with EDF (Electricité de France). His work, called ‘Analysis of very large multivariate time series using unsupervised learning methods in order to detect anomalies and support predictive maintenance’, focused on the analysis of very large volumes of time series data in the energy sector, collected from nuclear power plant reactors, hydroelectric dams, etc. Using new unsupervised learning methods, Paul created algorithms capable of detecting anomalies and facilitating their prediction. These results are available to the international scientific community. His work has been awarded 4 prestigious prizes and has led to the filing of national and European patents.