Sensors in Self-Driving Vehicles: A Detailed Literature Review and New Trends
Autonomous vehicles rely on complex sensing systems to perceive their environment and ensure safe operation. This review analyses the main sensor technologies used in self-driving vehicles, including cameras, LiDAR, radar, ultrasonic sensors and GNSS/IMU-based localisation systems. A core set of 40 primary research articles was systematically analysed to compare the capabilities, limitations and integration challenges of sensing technologies used in autonomous vehicles. In addition to these primary studies, further references were included to provide background information and describe emerging developments in autonomous sensing systems. The review shows that no single sensor technology can provide reliable perception under all environmental conditions. Camera systems offer rich visual information but are sensitive to lighting and weather conditions, while LiDAR provides highly accurate three-dimensional geometry but suffers from signal attenuation in rain and fog. Radar sensors demonstrate superior robustness in adverse weather and enable direct velocity measurement, although their spatial resolution remains limited compared to optical sensors. As a result, modern autonomous vehicles rely on multi-sensor fusion architectures that combine complementary sensing modalities to improve reliability and safety. The analysis also identifies several key research gaps in the current literature. In particular, there is a lack of systematic evaluation of trade-offs between sensor performance, computational requirements and vehicle energy consumption. Furthermore, the safety certification of artificial intelligence-based perception systems and the integration of emerging technologies such as FMCW LiDAR and terahertz radar remain open research challenges. Overall, the results suggest that the future of autonomous vehicle perception will depend not only on improvements in individual sensors but also on robust sensor fusion architectures, safety-certified AI models and energy-efficient sensor processing platforms. These findings provide guidance for researchers and engineers developing next-generation sensing systems for autonomous driving.