This paper examines the potential effectiveness of the following three precollision system (PCS) algorithms: 1) forward collision warning only; 2) forward collision warning and precrash brake assist; and 3) forward collision warning, precrash brake assist, and autonomous precrash brake. Real-world rear-end crashes were extracted from a nationally representative sample of collisions in the United States. A sample of 1396 collisions, corresponding to 1.1 million crashes, were computationally simulated as if they occurred, with the driver operating a precollision-system-equipped vehicle. A probability-based framework was developed to account for the variable driver reaction to the warning system. As more components were added to the algorithms, greater benefits were realized. The results indicate that the exemplar PCS investigated in this paper could reduce the severity (i.e., ΔV) of the collision between 14% and 34%. The number of moderately to fatally injured drivers who wore their seat belts could have been reduced by 29% to 50%. These collision-mitigating algorithms could have prevented 3.2% to 7.7% of rear-end collisions. This paper shows the dramatic reductions in serious and fatal injuries that a PCS, which is one of the first intelligent vehicle technologies to be deployed in production cars, can bring to highway safety when available throughout the fleet. This paper also presents the framework of an innovative safety benefits methodology that, when adapted to other emerging active safety technologies, can be employed to estimate potential reductions in the frequency and severity of highway crashes.
Objective: The objective of active safety systems is to prevent or mitigate collisions. A critical component in the design of active safety systems is the identification of the target population for a proposed system. The target population for an active safety system is that set of crashes that a proposed system could prevent or mitigate. Target crashes have scenarios in which the sensors and algorithms would likely activate. For example, the rear-end crash scenario, where the front of one vehicle contacts another vehicle traveling in the same direction and in the same lane as the striking vehicle, is one scenario for which forward collision warning (FCW) would be most effective in mitigating or preventing. This article presents a novel set of precrash scenarios based on coded variables from NHTSA's nationally representative crash databases in the United States. Methods: Using 4 databases (National Automotive Sampling System–General Estimates System [NASS-GES], NASS Crashworthiness Data System [NASS-CDS], Fatality Analysis Reporting System [FARS], and National Motor Vehicle Crash Causation Survey [NMVCCS]) the scenarios developed in this study can be used to quantify the number of police-reported crashes, seriously injured occupants, and fatalities that are applicable to proposed active safety systems. In this article, we use the precrash scenarios to identify the target populations for FCW, pedestrian crash avoidance systems (PCAS), lane departure warning (LDW), and vehicle-to-vehicle (V2V) or vehicle-to-infrastructure (V2I) systems. Crash scenarios were derived using precrash variables (critical event, accident type, precrash movement) present in all 4 data sources. Results and Conclusions: This study found that these active safety systems could potentially mitigate approximately 1 in 5 of all severity and serious injury crashes in the United States and 26 percent of fatal crashes. Annually, this corresponds to 1.2 million all severity, 14,353 serious injury (MAIS 3+), and 7412 fatal crashes. In addition, we provide the source code for the crash scenarios as an appendix (see online supplement) to this article so that researchers can use the crash scenarios in future research.