Simulation and Emulationof Water spray for Validation of Optical Sensors (SEVVOS)

This research investigated visibility degradation caused by vehicle-generated water sprayon wet surfaces, using experimental tests, simulations, and data analysis to examine spraydynamics and their effects on camera and sensor performance.Dynamic tests faced challenges with automated contrast analysis due to insufficientresolution, lack of camera calibration, and poor lighting. Targets were too small inimages, and low contrast, even without spray, prevented reliable detection. Similar issuesaffected static tests, although higher light levels enabled more consistent results. Highbeamheadlights worsened contrast degradation by illuminating spray particles. Thesefindings emphasized the importance of proper calibration, resolution, and lighting foraccurate data collection.Outdoor tests on AstaZero test tracks showed that water depth and vehicle speedsignificantly influence spray and visibility. Deeper water (e.g., 9–10 mm) caused greatercontrast degradation than shallower water (e.g., 4–5 mm), while higher speeds amplifiedspray effects, particularly in shallow water. Variations in light conditions affected theresults, with clearer patterns emerging under stable lighting.Tyre rig tests provided detailed measurements of aerosol and water spray properties, suchas droplet size, density, and distribution. Smaller droplets (mode below 50 μm) formednear the tyre surface, while larger droplets developed downstream due to coalescence andaerodynamic forces. Higher tyre speeds and more water increased spray density andcontrast degradation. In deeper water, contrast degradation was more uniform, withnarrower ranges between maximum and minimum values.Simulations revealed key mechanisms of spray generation and propagation. Water filmdepths as low as 100 μm produced spray through capillary adhesion, with dropletsinteracting with vehicle components and airflow. Larger droplets returned to the groundquickly, while smaller droplets remained suspended, affecting visibility. Data collectedunder naturalistic conditions validated these findings and provided insights into realworldvisibility challenges.This research highlights the critical role of water depth, vehicle speed, and spraydynamics in visibility degradation. It underscores the need for improved measuringmethods, lighting, and testing protocols to enhance automated analysis and sensorperformance, especially for autonomous vehicle systems in adverse weather conditions.