Abstract:
Intersection crashes are among the most frequent and lethal crash modes in the United States. Accounting for over one-third of all intersection crashes, straight crossing...Show MoreMetadata
Abstract:
Intersection crashes are among the most frequent and lethal crash modes in the United States. Accounting for over one-third of all intersection crashes, straight crossing path (SCP) crashes are the most common intersection crash mode. Intersection Advanced Driver Assistance Systems (I-ADAS) have the potential to prevent SCP crashes by detecting imminent collisions and either alerting the driver and/or taking autonomous crash avoidance action. The objective of this study was to estimate how many SCP intersection crashes could be potentially prevented in the U.S. if every vehicle was equipped with I-ADAS. Three steps were performed in this study. First, a simulation case set was generated from 459 real world SCP intersection crashes collected as part of NHTSA's National Motor Vehicle Crash Causation Survey (NMVCCS) database. Second, the pre-crash kinematics of each vehicle was reconstructed using information from the crash investigation, pre-crash driver models, and reconstructed impact speeds. Third, the crashes were simulated as if both vehicles had been equipped with I-ADAS. Three critical time-to-collision (TTC) thresholds were evaluated in this study, including 2.0, 2.5, and 3.0 seconds. The model predicted that 19% to 35% of all SCP crashes have the potential to be prevented if all vehicles in the U.S. were equipped with I-ADAS. Nearly twice as many crashes were predicted to be prevented if a TTC threshold of 3.0 s was used rather than 2.0 s. When at least one of the vehicles stopped prior to entering the intersection, the model estimated that 24% to 49% of crashes have the potential to be prevented by I-ADAS. In contrast, when neither vehicle stopped, the model estimates that 13% to 17% of crashes could potentially be prevented. It is important to note that the model makes several assumptions that represent a “best case scenario” for I-ADAS. These results have important implications for designers, consumers, and regulatory agencies.
Published in: 2016 IEEE Intelligent Vehicles Symposium (IV)
Date of Conference: 19-22 June 2016
Date Added to IEEE Xplore: 08 August 2016
ISBN Information: