Lane Keeping Support (LKS) is an advanced driver assistance system (ADAS) technology intended to prevent a vehicle from drifting out of its travel lane. To assess the potential for LKS to reduce real-world crashes where the driver drifts out of their travel lane, test track performance was compared with the real-world crash data.
Five light vehicles equipped with LKS were evaluated on the test track using Lane Keeping Assist (LKA) test methods contained within the Euro NCAP Test Protocol - Lane Support Systems. Specifically, the procedures to evaluate a vehicle’s response to an imminent departure over a solid white line were used; tests to evaluate LKS system response to an unmarked road edge were not performed. These tests identified performance differences between the vehicles, and were somewhat dependent on the lateral velocity used during test conduct.
Results from these tests were compared to relevant fatal crashes in the National Motor Vehicle Crash Causation Survey (NMVCCS) survey conducted by the National Highway Traffic Safety Administration from 2005-2007, and the agency’s new Crash Investigation Sampling System (CISS). A review of the fatal 2005 – 2007 NMVCCS and 2017 CISS lane/roadway departure cases was performed to classify the shoulder type present on the side of the roadway from which the subject vehicle first departed from, and to estimate the shoulder width just after the departure, where applicable. The objective of this effort was to estimate whether LKS interventions could have potentially amended the real-world pre-crash path of the subject vehicle in the vicinity of the lane departure, given the system performance observed on the test track.
When the test track performance of the vehicles was considered in the context of the road shoulder widths and road/lane/shoulder characteristics present in the 43 fatal NMVCCS and 50 CISS crashes analyzed for this paper, estimating whether LKS could have affected the crash outcome was found to depend on a number of factors. From an input perspective, the lateral velocity of the vehicle as it is directed toward the boundary of the lane, and whether that boundary is comprised of a clearly defined painted line or simply a pavement edge, has the potential to affect whether an LKS intervention can even be expected.
Even if the input conditions are such that a vehicle’s LKS activation criteria are satisfied, then the ability of the system to effectively address the pre-crash scenario is relevant, yet can depend on a number of factors. The amount of lateral deviation before or beyond the lane line and/or road edge, and the implications of it being too large, are important considerations. In the case of a right-side departure away from the travel lane, excessive lateral deviation may result in at least part of the vehicle leaving the paved roadway. Similarly, left-side departures with excessive lateral deviation have the potential to increase the risk of a head-on crash.