Automobile safety can be improved by anticipating a crash before it occurs and thereby providing additional time to deploy safety technologies. This requires an accurate, fast and robust pre-crash sensor that measures telemetry, discriminates between classes of objects over a range of conditions, and has sufficient range and area of coverage surrounding the vehicle. The sensor must be combined with an algorithm that integrates data to identify threat levels. No one sensor provides adequate information to meet these diverse and demanding requirements. However the requirements can be met with an optimal combination of multiple types of sensors. Previous work considered criteria for evaluating various sensors to find an optimal combination. This work presents test methods and results for selected sensors proposed for use in a pre-crash detection system. The test methods include static and dynamic telemetry testing to identify the range, accuracy, reliability and operating conditions for each sensor. Each sensor is evaluated for its ability to discriminate between classes of objects. The tests are applied to ultrasonic, laser range finder and radar sensors. These sensors were selected because they provide the maximum information, cover a broad range and region and are commercially viable in passenger vehicles.


Mechanical Engineering

Number of Pages


Publisher statement

Reprinted with permission from SAE. Paper Number 06AE-19.



URL: https://digitalcommons.calpoly.edu/meng_fac/26