This research aims primarily to estimate the lifespan of one of the key components of the traction chain, namely the windings of the traction electric motor. The degradation of the windings, particularly the insulation of the copper wire, is explained by the shape of the signals provided by the inverter. The models developed will rely mainly on a hybrid AI/physics approach, which will provide insights into the causes of the degradation (temperature, insulation aging, the appearance of partial discharges, etc.).