In June 2016, the National Highway Traffic Safety Administration released a field study on the use of collision avoidance systems in heavy vehicles. The report will help government agencies and transportation companies wanting to stay ahead of the curve in the implementation of more advanced safety features in the trucking industry.
The study started by road-testing in cooperation with a number of willing trucking companies. All who were paid a small fee by the NHTSA. The test used four different truck models, all models from 2013 or later. Additionally, the test used two separate vehicle avoidance systems: the Meritor WABCO OnGuard and the Bendix Wingman Advanced.
The study then sought to classify collision avoidance system activations and alerts into a variety of event types, such as following distance alerts, lane departure warnings, impact alerts, stationary object alerts, and automatic emergency braking. The study collected data from 110,000 hours of driving, 3,245,000 miles traveled, and over 885,000 activations and alerts. From this data, the NHTSA study pulled a representative sample of 6,000 activations and studied video from each.
The study had a number of objectives both short term and long term.
The field study did result in a number of interesting findings. Strikingly, while the collision avoidance systems performed very reliability for most activations and alerts, the study found a high number of false stationary object alerts - nearly 98% of the stationary object alerts happened without justification. Hopefully this leads to CAS manufacturers addressing the issue in future versions before teaching those systems to activate control of the vehicle upon receiving a stationary object alert.
The study also failed to find any key modifications to driver behavior over time as a result of exposure to the collision avoidance system activations and alerts. While the study tracked some changes to following distance over time, statistical analysis deemed these changes insignificant. Researchers observed no changes to driving speed, reaction time, or average maximum deceleration.
In the report’s conclusion, the researchers sought to highlight potential safety benefits of collision avoidance systems while noting both the limitations of their study and the still large potential for improvement in these systems. In the 6,000 sampled activation and alerts, the study found zero actual crashes. As part of the study, five involved fleet safety managers were asked if they would recommend the technology to colleagues in other companies. All five said yes.