2 physicists journey: from condensed matter physics to rolling stock maintenance at SNCF
par
,Salle des Séminaires
Guido Intronati and Héloïse Nonne are two former condensed matter physicists and they now work at SNCF, managing a team of 25 data scientists and engineers. They will explain how they chose to switch from fundamental physics to operational work in the railroad industry, what skills are common to both worlds, and what kind of projects they lead.
To illustrate their work in more detail, they will present a recent project on the maintenance supply chain. In order to better manage spare parts stocks, their teams developed an AI algorithm that helps supply chain buyers decide to approve or reject buy orders. Based on historical data, the algorithm calculates a score for overstock risk associated to each new order, along with a graph providing additional information that explains the reasons of the prediction. The buyer can use this graph to interpret the prediction, detect risky orders and investigate in more detail before approving an order. The algorithm is now deployed in production for testing in real-life conditions.