Event Data Recorder Duration Study [Appendix to a Report to Congress. Report No. DOT HS 813 082B]

Chen, Rong Jackey; Tatem, Whitney; Gabler, H. Clay · 2022 · ROSA P / United States. Department of Transportation. National Highway Traffic Safety Administration

archive: archived pipeline: cataloged verified

Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)

Summary

This study, conducted by researchers at Virginia Tech for the National Highway Traffic Safety Administration (NHTSA), addresses the limitation of current Event Data Recorders (EDRs), which typically capture only five seconds of pre-crash data. The research aims to determine the recording duration necessary to fully investigate crash causation and understand driver pre-crash behavior. The motivation stems from the observation that five seconds may be insufficient to capture critical driver inputs, such as braking or steering, particularly in scenarios like car-following events where time-to-collision often exceeds this window. Understanding these durations is critical for evaluating emerging active safety systems and reconstructing crash sequences. The research was executed in two phases. Phase 1 analyzed EDR data from the National Automotive Sampling System/Crashworthiness Data System (NASS/CDS) and the National Motor Vehicle Crash Causation Study (NMVCCS) to evaluate the frequency with which current EDRs failed to capture driver actions in rear-end, intersection, and road departure crashes. Researchers calculated the time between driver inputs (brake, steering, accelerator) and impact to assess if the five-second window was adequate. Phase 2 utilized naturalistic driving data from the 100-Car Naturalistic Driving Study and the Second Strategic Highway Research Program (SHRP-2) NDS to analyze the complete duration of driver actions during normal driving conditions, including car following, lane departure, and intersection traversal. This phase focused on determining the actual time required for maneuvers like approaching intersections or recovering from lane drifts, providing a baseline for necessary recording lengths beyond what crash-only EDR data could reveal. The findings indicate that the standard five-second EDR duration is frequently insufficient for capturing the full sequence of pre-crash events. In Phase 1, analysis of NASS/CDS data revealed that a significant portion of crashes involved driver inputs initiated before the five-second recording window began, particularly in rear-end and intersection scenarios. For instance, many drivers initiated braking or steering maneuvers more than five seconds prior to impact, meaning the EDR recorded only the tail end of their avoidance attempts. Phase 2 results from naturalistic driving studies demonstrated that typical maneuvers, such as approaching and traversing intersections or recovering from lane departures, often require durations exceeding five seconds. For example, intersection traversal times varied significantly based on intersection size and control type, with many events requiring longer recording windows to capture the complete approach and traversal phases. The significance of this study lies in its recommendation for extending EDR recording durations to better support crash causation analysis. The authors conclude that current EDRs cannot provide insight into the specific duration beyond five seconds needed to capture crash causation because they lack the data to begin with. However, the naturalistic driving data suggests that longer recording times would enable a more comprehensive understanding of driver behavior, such as car-following habits and intersection navigation. This extended data would facilitate the evaluation of advanced driver assistance systems and improve the accuracy of crash reconstruction, ultimately supporting safer vehicle design and traffic safety policies.

Key finding

The standard five-second pre-crash recording duration of most EDRs is often insufficient to capture the initiation of driver errors or avoidance actions, particularly in rear-end and intersection crashes, necessitating longer recording windows to fully characterize crash causation.

Methodology

naturalistic

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• naturalistic crash near crash
• braking response
• perception reaction time
• crash reconstruction hf
• incidence prevalence
• exposure measurement

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: crash risk outcomes, behavioral performance data
• Methodological Resource: dataset resource