The objective of this study was to prioritize the variables that could be transmitted with an ACN (Automatic Crash Notification) signal. The main purpose of transmitting these variables is to assist in early identification of those occupants with time critical injuries. For the purposes of this study, all MAIS 3+ injuries were classified as time critical. The basis for prioritizing crash variables was based on their ability to identify MAIS 3+ injured occupants in the National Automotive Sampling System- Crashworthiness Data System (NASS/CDS) dataset.
In this study, multivariate models to represent crash events were developed based on historical crash data from the years 1997-2003. The analysis established a relationship between crash attributes and crash outcomes for all passenger vehicles in the database.
The resulting analysis provided a ranking of crash variables in order of importance. Crash severity (Delta-V) was found to be the most important variable for all planar crash directions. The addition of other crash variables improved the accuracy of the injury prediction algorithm.
For frontal crashes important secondary crash variables include: 3-point belt usage, multi-impact crashes, occupant age and the presence of more than 6” of intrusion. For near-side crashes, the most important secondary variables were occupant age, narrow object crashes, and the presence of intrusion. For far side crashes, the most important secondary crash variables were 3-point belt usage and the occurrence of a narrow object crash. Rollover was found to be a high risk event that predicted high injury risk independent of Delta-V if 3-point belts were unused.
The paper will show the relative importance of the crash and occupant variables by crash direction.