With the development of advanced driving assistance systems (ADAS), the integration of camera systems behind automotive glazing becomes more and more common. No standardization of specifications exists for glass used with ADAS. Understanding the right level of quality to ensure a maximum system performance is therefore key.We developed an in-house simulation tool based on stochastic ray tracing to render its impact on ADAS camera images in a more accurate way. In this simulation tool, the windshield and the camera lens are modelled separately, which paves the way towards modelling the entire optical system and understanding its impact on ADAS performances. Object detection algorithms can then be run on the simulated images to assess the impact of the windshield and design ADAS-compliant systems without spending thousands of hours on road