Crash safety concerns for out-of-position occupant postures: A look toward safety in highly automated vehicles

Objective: Highly automated vehicle occupants will all be passengers and may be free to ride while in postures for which existing occupant safety systems such as seat belts and airbags were not originally designed. These occupants could therefore face increased risk of injury when a crash occurs. Given that current vehicles are capable of supporting a variety of occupant postures outside of the normal design position, such as reclined or turned passengers, an evaluation of current field data was performed to better understand the risks of being out of position.

Methods: We investigated the frequency, demographics, and injury outcomes for out-of-position occupants using NASS-CDS. A matched analysis was performed to compare injury outcomes for out-of-position passengers with in-position drivers involved in similar crashes. Finally, case studies for out-of-position occupants were examined in the Crash Injury Research (CIREN) database.

Results: Only 0.5% of occupants in NASS-CDS with a coded posture were out of position at the time of crash. Of the out-of-position occupants, being turned or seated sideways was almost as likely as being reclined. Out-of-position occupants were younger and less likely to be belted than their in-position counterparts. Analysis of the injury data indicated a trend that being out of position was associated with an elevated risk for serious injury. However, the number of out-of-position occupants was too small to provide a definitive or statistically significant conclusion on injury outcome.

Conclusion: Though highly automated vehicles may eventually reduce the number of crashes and traffic fatalities in the future, there will be a transition period when these vehicles remain at risk from collisions with human-driven vehicles. These crashes could cause higher than anticipated rates of injury if occupants are less likely to be belted or tend to be in positions for which restraints are not optimized. This study highlights the need for future research on occupant response and countermeasure design for out-of-position occupants.