Crash causation studies that stretch back to the 1960s have consistently reported human (primarily driver) errors as the cause of the overwhelming proportion of roadway collisions. Application of driver alert and active collision avoidance technologies may begin to affect drivers’ pre-collision actions and their resultant success in crash avoidance or injury mitigation when a crash does occur. With the introduction of connected vehicle (V2X) technologies, vehicle-to-vehicle communications will better inform drivers to take avoidance actions or in some cases even automatically control the longitudinal and lateral dynamics of the vehicle so as to avoid collisions and mitigate injury should a crash occur regardless. Further, the great promise of automated 1 vehicle systems is the elimination of most driver observational, judgement, and controls actuation errors, thus resulting in collision avoidance or in injury mitigation should a crash occur. Safety researchers anticipate that these systems now emerging as new safety technologies, or currently in advanced research stages will provide significant public health benefits but may not prove to be one hundred percent effective in collision avoidance.

The sensor inputs, controls algorithms, and driver alerts and/or vehicle systems actuations that may be commanded by Advanced Driver Assist systems or by various levels or automated driving systems are engineered parameters and will be well understood at introduction of the systems into the stream of commerce. However, as vehicles equipped with advanced collision avoidance technologies and automated driving systems are anticipated to continue to be involved in some crashes, it is essential that safety researchers, engineers, and regulators are able to develop a complete understanding of those collisions that continue to occur and why such collisions did occur. Conventional accident reconstruction techniques will be insufficient to the task of understanding pre-crash conditions, changes in conditions observed prior to impact, and post- impact events. Therefore, research demands for data related to prevailing conditions, conditional awareness, and post-crash data availability are critical to development of understanding of crash causation and further refinement of safety systems through study of customer use experiences. This paper introduces some criteria for selection of pre- crash, collision, and post-crash related data that may be of use in understanding crash causation in advanced crash avoidance platforms and in engineering refinements in second and subsequent generations of advanced collision avoidance technologies including automated driving system equipped vehicles.