Performance Evaluation for a Muddy Water Contamination on Autonomous Sensors

Abstract

Autonomous driving systems depend heavily on advanced recognition sensors, such as light detection and ranging, radar, cameras, and ultrasonic sensors, to navigate dynamic environments safely and effectively. However, these sensors are vulnerable to environmental contamination, particularly from muddy water, which can significantly degrade their optical performance. This letter presents a standardized method for evaluating the impact of muddy water contamination on recognition sensor cover with a focus on assessing transmittance reduction. An automated injection system was employed to apply varying levels of contamination to polycarbonate windows, and transmittance was estimated based on a pulsed laser module. The proposed method demonstrated that practical applicability was achieved by comparison with real-world contamination during test driving and laboratory findings. This research will provide a reliable, standardized approach to quantify the effects of various contaminations on sensor performance, also contributing to the development of more robust autonomous driving technologies capable of operating in harsh environmental conditions.