​Transdev, the global leader in mobility services, turned to Leti, a CEA Tech institute, to improve autonomous vehicle reliability and safety. Leti is running tests to characterize and evaluate commercially-available laser sensors to assist Transdev in selecting the most suitable sensors for the target applications. Characterizing the sensors' behavior is crucial to developing autonomous vehicles equipped with algorithms that are safe in the widest possible range of environmental conditions.

Leti drew on its know-how in the fields of sensor development and embedded data-fusion algorithms to characterize the efficiency and robustness of various commercially-available LiDAR sensors. Leti researchers evaluated the sensors' behavior and responses in real-world conditions, including exposure to objects with fluctuating reflectivity like tires and road signs. They also tested the sensors in different weather, exposing them to varying degrees of sunlight, fog, snow, rain, and glare.

The project resulted in a performance evaluation of the sensors as well as a list of objective criteria and parameters that can be used to evaluate other commercially-available LiDAR systems.