Abstract
Research and/or Engineering Questions/Objective
Event data recorders (EDRs) record valuable data in estimating the occupant injury severity after a crash. Advanced automatic collision notification (AACN) with the use of EDR data will determine the potential extent of injuries to those involved in motor vehicle accidents. The objective of this study is to find essential EDR data elements (EDR risk factors) and their thresholds to be used for injury prediction.
Methodology
Selected accident types were single vehicle frontal collisions and vehicle‐to‐vehicle frontal collisions. As a first step, frontal crashes were chosen and rear‐end collisions and side impacts were excluded in this study. Accident vehicles selected for the analysis were passenger cars equipped with an EDR and having a comprehensive automobile insurance policy, in which the front airbags were deployed. Injury severities of drivers and of occupants were collected from the accident database of insurance companies and defined based on the abbreviated injury scale (AIS), the maximum abbreviated injury scale (MAIS) and the injury severity score (ISS). These scales are internationally used for ranking injury severity in motor vehicle crashes. For the EDR data retrieved from the accident vehicles, four items (collision speed (V0), velocity change during collision or delta V (V), seating position, and seat belt usage) were analyzed. For the occupant information, five items (occupant age, MAIS, ISS, impact configuration, and vehicle type) were selected.
Results
The maximum delta V (Max V) and the maximum abbreviated injury scale (MAIS) were generally in proportion to each other. However, there were a few cases in which the occupant injury was serious although the Max V was small. There were also several cases with slight injury under higher Max V. There will be potential errors of overestimation (overtriage) at higher Max V and underestimation (undertriage) at lower Max V when predicting the level of injury based upon the Max V alone. Accordingly, combinations of EDR risk factors and their thresholds for predicting the serious injury were studied in order to avoid the undertriage and to minimize the overtriage. As a result, the collision speed (V0) and the maximum velocity change within 50 milliseconds during collision (Max V 50ms) showed an important role on the injury severity. Finally, three EDR risk factors (Max V, V0, Max V 50ms) and their thresholds were proposed for the injury estimation.
Limitations of this study
Frontal collisions were chosen in this study. Further study in other crash configurations such as side impact, rear‐end collision and rollover will encourage the improvement of the injury estimation using EDR data. This study is preliminary and will need further confirmation using a larger database to adequately evaluate the practical use of EDR data for AACN.
What does the paper offer that is new in the field including in comparison to other work by the authors?
In the previous study, relatively small size of the data source were analyzed as a pilot study in order to obtain basic information in injury estimation using EDR data. The current study, by increasing the number of accident vehicles equipped with EDRs, shows that the information recorded on an EDR has the potential for assisting emergency medical service (EMS) in estimating the level of injury expected for occupants.
Conclusions In the injury estimation with the use of EDR data, it seems possible to avoid the undertriage and to minimize the overtriage as much as possible. However, this method will need further validation in reliability and accuracy by using larger database for the practical use into AACN. This study also shows that the information on the occupant age seemed to be very important in the injury prediction. Further, in a frontal collision, even at higher Delta V, the vehicle safety performance for the restrained occupant seemed to have worked effectively in decreasing the injury severity.
Key Words : Safety; Automatic collision notification; Injury prediction; Advanced automatic collision notification