The Perception System of Intelligent Ground Vehicles in All Weather Conditions: A Systematic Literature Review

Area of Science:

• Automotive Engineering
• Sensor Technology
• Artificial Intelligence in Transportation

Background:

• Driving perception is critical for safety, yet compromised by weather (snow, fog, rain), causing global accidents.
• Existing automated driving systems require robust perception capabilities across all environmental conditions.

Purpose of the Study:

• To survey sensing technologies crucial for detecting the driving environment and obstacles in diverse weather conditions.
• To analyze the performance of current automated driving applications concerning all-weather functionality.

Main Methods:

• Review of historical milestones in driving perception technology.
• Examination of state-of-the-art automated driving applications, focusing on all-weather capabilities.
• In-depth study of key sensor technologies: radar, lidar, ultrasonic, camera, and far-infrared sensors.

Main Results:

• Comparison of current sensor performance against expected states using spider charts.
• Identification of performance gaps in existing perception systems under adverse weather.

Conclusions:

• A sensor fusion perspective is proposed to address perception system limitations.
• Sensor fusion is identified as a key strategy to increase the robustness and reliability of automated driving systems in all weather conditions.