It has been reported that on Friday, May 11, 2018, a Tesla Model S sedan crashed into the back of stopped fire truck at a speed of 60 mph (100 km/h) in South Jordan Utah. The driver of the Tesla, or its semi-autonomous system, did not apply any braking prior to the impact and the driver only suffered a broken right ankle.
The above photo of the crashed Tesla was released by South Jordan police. This photo shows that crush to the vehicle was mainly above its bumper in a typical “under-ride” manner that would not be unexpected given the typical higher rear bumper of a large truck. The front wheels of the Tesla do not appear to have been displaced rearward and the A-pillars also appear to be in their original positions. All these facts support the conclusion that the Tesla’s crush does not exhibit the type of expected energy dissipation that would be comparable to a change-in-velocity (Delta-V) anywhere close to 60 mph (100 km/h).
Ever since the published research paper by Kenneth Campbell in 1974 comparing change-in-speed and vehicle crush there has been a lot of “crush measuring” in the field by various collision investigators/reconstructionists. In fact, computer reconstruction programs such as CRASH were developed for the U.S. National Highway Traffic Safety Administration (NHTSA) for the specific purpose of objectively estimating the severity of a crash through the measurement of vehicle crush. So those who are familiar with those procedures know that the stiffness of most light-duty vehicles is much less in the vertical zone above the bumper than it is a the bumper level. Thus the crush shown in the Tesla would be much less had it involved contact with its bumper where the stiff structure is located.
It is true that under-ride collisions such as that shown in the above Tesla photo result in an under-estimation of the speed loss because the impact force is not horizontal but there is a vertical component which tends to lift the taller struck vehicle (i.e. the fire truck) while pushing the striking vehicle into the pavement. Furthermore some speed will be lost from post-impact motion of the vehicles, yet it is unlikely that the post-impact travel distances would be substantial in that the fire truck would act like a massive, movable barrier due to the large difference in the masses of each vehicle. While these factors are known they cannot explain the very large difference in the visible crush and the reported speed loss of close to 60 mph.
The unusually large interest in this crash is due to the possibility that its auto-pilot mode may have been activated and therefore this could suggest that the system failed to apply braking when it would be expected to do so. The present information indicates that it is unknown at this time whether that system was indeed activated.
From a collision reconstruction viewpoint, the use of event data from downloading of data from a recorder is becoming increasing common such that many investigators and analysts are becoming exclusively dependent on such data when completing their conclusions about how a collision occurred. Whatever the eventual circumstances in the present collision, it demonstrates the importance of investigators/analysts knowing their physical evidence and being able to use that knowledge of the physical evidence to independently detect possible discrepancies that may exist in downloaded event data. This comment is not just with respect to the physical evidence visible in vehicle damage, but relates to all the evidence on the roadway and with respect to the expected performance of the human behind the wheel.
Recent court appearances in the Province of Ontario demonstrate that many courts appear oblivious the wide range of knowledge that a collision reconstructionist must have, and must evaluate in order to narrow possible causes to one or select few.