A Cost-Effective and Climate-Resilient Air Pressure System for Rain Effect Reduction on Automated Vehicle Cameras

Recent advances in automated vehicles have focused on improving perception performance under adverse weather conditions; however, research on physical hardware solutions remains limited, despite their importance for perception critical applications such as vehicle platooning. Existing approaches, such as hydrophilic or hydrophobic lenses and sprays, provide only partial mitigation, while industrial protection systems imply high cost and they do not enable scalability for automotive deployment. To address these limitations, this paper presents a cost-effective hardware solution for rainy conditions, designed to be compatible with multiple cameras simultaneously. Beyond its technical contribution, the proposed solution supports sustainability goals in transportation systems. By enabling compatibility with existing camera-based sensing platforms, the system extends the operational reliability of automated vehicles without requiring additional high-cost sensors or hardware replacements. This approach reduces resource consumption, supports modular upgrades, and promotes more cost-efficient deployment of automated vehicle technologies, particularly in challenging weather conditions where system failures would otherwise lead to inefficiencies and increased emissions. The proposed system was able to increase pedestrian detection accuracy of a Deep Learning model from 8.3% to 41.6%.

💡 Research Summary

The paper addresses a critical gap in autonomous‑vehicle perception: the degradation of camera‑based sensing under rain. While many studies focus on algorithmic compensation or expensive industrial hardware, the authors propose a low‑cost, physically‑based Air Pressure System (APS) that creates a directed air curtain over each camera lens to physically dislodge water droplets. The system is built from off‑the‑shelf components—a 220 V AC‑compatible air compressor capable of 338 km/h airflow and 12 m³/min volume, weather‑resistant tubing, quick‑connect fittings, Y‑junctions, and custom‑fabricated nozzles that attach at a 90° angle to the lens. All parts together cost less than €100, making the solution scalable to multiple cameras without major redesign.

The APS was installed on the JKU‑ITS research vehicle and evaluated using the central forward‑facing camera. Experiments were conducted in real‑world rainy conditions, recording 100 seconds of video both with and without the APS. Pedestrian detection performance was measured using YOLOv4‑tiny, chosen for its balance of speed and accuracy on embedded hardware. Two evaluation methods were employed: a qualitative check that a pedestrian appears in at least three consecutive frames, and a quantitative metric calculating the percentage of frames in which a pedestrian is correctly detected over ten‑second intervals.

Results show a dramatic improvement. In clear weather the baseline detection rate is 100 %. Under rain without APS the detection rate collapses to 8.3 %. When the APS is activated, the detection rate rises to 41.6 %, representing a more than five‑fold increase. Qualitative frame samples confirm that the air curtain removes water droplets from the lens, preserving image clarity and enabling continuous detection.

Beyond performance, the authors argue that the APS aligns with sustainability goals. Conventional industrial solutions (heated enclosures, specialized coatings) often cost several thousand euros per unit, require dedicated power supplies, and demand frequent maintenance, all of which increase material consumption and energy use. By contrast, the APS leverages existing vehicle power, uses inexpensive components, and can be retrofitted to existing camera mounts, reducing both upfront capital expenditures and long‑term operational costs.

The paper also acknowledges limitations. The experimental validation is limited to a single camera and a specific rain intensity; broader testing across varying precipitation rates, wind directions, and vehicle speeds is needed to confirm generalizability. Potential side effects such as acoustic noise, vibration, and aerodynamic drag introduced by the airflow were not quantified. Energy consumption of the compressor and its impact on vehicle battery life were not reported, leaving open questions about the system’s efficiency during extended missions. Finally, the fixed‑angle nozzle design may not be optimal for cameras mounted at different orientations, suggesting a need for adjustable or adaptive nozzle geometries.

Future work is outlined to address these gaps: (1) extensive field trials under diverse weather conditions and longer durations, (2) optimization of nozzle shape and airflow to minimize noise and power draw, (3) integration of a power‑management module to monitor and limit compressor load, and (4) extension of the modular architecture to other external sensors such as LiDAR and radar.

In summary, the proposed APS offers a pragmatic, cost‑effective hardware approach to mitigate rain‑induced image degradation on vehicle cameras. By physically removing water droplets, it substantially improves detection performance, supports multi‑camera scalability, and contributes to reduced material and energy consumption, making it a promising candidate for widespread adoption in climate‑resilient autonomous‑vehicle platforms.