A crash happens in the time it takes to read this sentence. The physical evidence — crush, momentum, and the vehicle's own data — reconstructs the speed, path, and sequence with or without a witness.
Start a conversation with our AI Research Concierge, already scoped to vehicle collisions. Pick a starting point, or describe your situation directly.
A vehicle collision converts kinetic energy into crush, and that conversion is quantifiable: the depth and shape of crush on each vehicle, applied to that vehicle's specific stiffness properties, yields an energy estimate that constrains impact speed, while momentum conservation applied to the pre- and post-impact paths constrains it independently. Where those two methods agree with each other and with the physical scene evidence — skid, yaw, and scuff marks, final rest positions — the speed and sequence are on solid ground. Modern vehicles add a third, often decisive input: the event data recorder, which can record pre-impact speed, braking, throttle, and steering input directly, turning a physics estimate into a measured value.
Each collision raises a distinct physics question, and separating them determines which evidence and analysis the case actually needs.
Conservation of momentum and crush-energy analysis establishing pre-impact speeds from post-impact positions and vehicle damage.
Skid, yaw, and scuff marks correlating to speed, braking effectiveness, and steering input in the moments before impact.
Tripped versus untripped rollover mechanics, critical sliding velocity, and the vehicle's stability factor at the point control was lost.
Tire tread separation, brake failure, or a steering or suspension defect that contributed to or caused the loss of control.
Seatbelt and airbag performance correlated to occupant movement inside the vehicle and the resulting injury causation.
Perception-reaction time, distraction indicators, and evasive-maneuver timing established from EDR data and scene evidence.
Vehicle collision work combines physical scene evidence, vehicle inspection, and recovered electronic data into a single, cross-checked reconstruction.
Vehicle collision cases routinely combine several of these:
Crush profiles, tire and brake condition, and the EDR module are the evidence. Repairing, salvaging, or releasing a vehicle before inspection can permanently erase the data needed to establish speed and cause.
It is generally treated as strong, objective evidence when the download is performed and documented correctly, because it records values the vehicle's own systems measured rather than a reconstructed estimate. It is not infallible — recording duration, parameter availability, and accuracy vary by manufacturer and model year, and a hard reset or severe power interruption can prevent a full download. It is best used alongside, not instead of, physical evidence.
Yes — this is the traditional core of the discipline. Crush-energy analysis, momentum conservation using pre- and post-impact vehicle paths, and skid or yaw mark analysis using measured drag factors can each independently estimate impact speed. Using more than one method and checking that they agree is standard practice and often produces a more defensible result than EDR data alone.
By separating the event into individual impacts and analyzing each one's momentum and crush evidence, then sequencing them using timing evidence — following-distance physics, brake-light and impact timing from any video, and the damage pattern each vehicle shows from being struck versus striking. The sequence usually matters as much as any individual speed, since it determines which driver had time to react to what.
An untripped rollover occurs from steering and lateral-friction forces alone — the vehicle rolls without first striking anything, typically during an aggressive evasive maneuver on pavement with a vehicle whose stability factor is marginal. A tripped rollover occurs when a lateral force from an external object — a curb, soft shoulder, or another vehicle — abruptly halts the tires' sliding and converts lateral momentum into a rolling motion. The distinction matters for both cause and, frequently, vehicle-design liability.
Both vehicles in their as-damaged condition before repair or salvage, the EDR data before further driving or a jump-start risks overwriting it, tire and brake condition, and the scene's skid, yaw, and debris evidence, which is often gone within days. Any dash-camera, surveillance, or 911-call audio should be requested promptly given typically short retention windows.
Technical briefings from our work in this area.
How severely a vehicle was struck and how an occupant was loaded are different questions with different evidence. Restraint hardware answers the second better than any severity number.
readTwo speed methods with almost no shared inputs. When they converge the result is strong; when they diverge, the divergence is usually telling you something about the collision.
readAn EDR is a restraint controller with a short memory, not a flight recorder. What it measured, what it derived, and what it never saw are three different things.
readTell us what happened. We will triage it and connect you with the right expert — usually within one business day.