A Statistical Analysis of Data from Heavy Vehicle Event Data Recorders

The most common trigger for event data collection in Heavy Vehicle¹ ECMs is a sudden decrease in the calculated vehicle speed. The calculated vehicle speed is a by-product of programmed calibrations and measured wheel speed data. In some cases, as is the case with Detroit Diesel ECMs, event data are recorded when the vehicle transitions from a driving state to a stopped state. Event data are reported with respect to time when the calculated vehicle speed change exceeds the preset threshold value or the first recorded 0 mph value. Because the data are not necessarily centered on the collision event itself, determination of impact speed and analysis of driver response can be problematic. A statistical evaluation of crash and non-crash related Heavy Vehicle Event Data Recorder (HVEDR) reports was conducted to identify specific measurable characteristics that can be used to identify the time of impact within reported event data. Statistically significant differences appeared between the data from crash-related reports and data from non crash-related reports, using both within- and between-subjects tests. Specifically, sudden changes in reported vehicle speed (ΔS) during the braking event were statistically significant predictors of the time of impact. By identifying the time of impact, EDR data can be related in time and space to identifiable physical evidence, and driver response time can be evaluated.