Who Crossed the Centerline?How Crash Reconstruction Challenged the Police Report
By Jodie Immell, BSME, ACTAR
The Case: An Impact Location That Didn’t Tell the Whole Story
Two vehicles approached each other on a curve, with one traveling northbound and the other southbound. The roadway curved right for the northbound driver. They collided in the northbound lane.
Police concluded that the southbound driver swerved left into the northbound vehicle’s lane. Our investigation examined whether the physical evidence and available vehicle data supported that sequence. The central question was who crossed the centerline before the drivers responded—not simply where the vehicles collided.

The Analysis: Reconstructing the Approach to Impact
The northbound vehicle had an event data recorder (EDR), often called a vehicle’s “black box.” The older southbound vehicle had no available EDR data.
The northbound vehicle’s recorded steering appeared generally straight, followed by three data points of right steering at 0.5-second intervals. Taken exactly as written, this would be 1.5 seconds of steering. However, a steering path calculated from those values did not follow the roadway when the end of that path was positioned at the impact location. At the vehicle’s speed shown in the data, following the curve required right steering for considerably longer.

We used several steps in our forensic accident reconstruction to figure out what really happened in the crash:
- Compare the calculated path with the roadway. We evaluated whether the recorded steering supported travel 100% within the northbound lane.
- Evaluate recording limitations. We examined how sampling intervals, timing, and steering resolution affected interpretation of the EDR data.
- Test potential steering paths. We evaluated a range of inputs consistent with those limitations, rather than one path.
- Assess speed and available time. We considered how vehicle speed affected travel distance and the opportunity for evasive steering.
- Compare the proposed movements with physical evidence. We evaluated vehicle damage, impact orientation, and roadway evidence for consistency.
Why the Reported Steering Values Required Interpretation
This vehicle’s steering data was recorded every half second. These samples gave us snapshots of steering, but left some uncertainty about what happened in between.
Three recording characteristics mattered:
- Timing of the final sample: Up to half a second could separate the last steering sample from the crash. Right steering could therefore have continued longer than the displayed data suggested.
- Timing of steering onset: Steering could begin between consecutive samples, placing its onset earlier than the first recorded change.
- Steering resolution: Values were reported in truncated 2.5-degree increments. A reported zero could represent a small steering input below the next increment.
These differences changed the calculated path during the five seconds leading up to the crash.
Understanding these limitations allowed us to evaluate the range of vehicle movements supported by the data. Each potential path also had to fit the physical evidence.
The Findings: Both Vehicles Steered Before Impact
Scroll through the three views, or select a numbered step.
Aligned Impact Orientation



Aligned Impact Orientation
The substantial angle between the vehicles at impact provided a critical clue. One vehicle steering into an otherwise lane-following vehicle could not reasonably explain the angle of damage between the vehicles as shown.
Damage overlaid on Simulation
The damage orientation supported evasive steering by both drivers.
Simulation Vehicle Impact Orientation
The damage orientation alone supported both vehicles in the southbound lane prior to impact.
The roadway geometry required sustained right steering for the northbound vehicle to follow the curve within its lane. However, the EDR showed only three samples of right steering immediately before impact, preceded by generally straight steering. Even considering the range of timing and resolution, the northbound vehicle did not have enough steering to follow the roadway. This data, combined with the placement of the curve relative to the impact, means that the northbound vehicle continued straight across the centerline before steering right for the curve. Remaining in the northbound lane would have required the driver to begin steering right substantially earlier. The EDR data alone, coupled with the geometry of the road indicated that the northbound vehicle entered the opposing lane before its final steering maneuver.
Two independent analyses show that the northbound driver crossed the centerline. Just before the crash, the northbound driver steered right, and the southbound driver steered left. Both vehicles consequently moved toward the northbound lane, where they collided. This sequence reversed the police report’s explanation of how the conflict began.
The Attorney Takeaway: Test the Sequence Behind the Impact
The lane containing the impact does not, by itself, establish which driver created the emergency. Evasive movements can place a collision in the lane originally occupied by the vehicle that crossed the centerline.
For attorneys and claim professionals, this case highlights three practical considerations:
- Evaluate EDR interpretation alongside the reported values. Sampling and resolution can materially affect the movements those values support.
- Test competing explanations against all available evidence. A proposed sequence must account for roadway geometry, speed, vehicle paths, and damage orientation.
- Use graphics to explain the sequence. Path comparisons and impact diagrams can help distinguish the initial lane departure from subsequent driver responses.
This approach supports litigation by identifying the evidence supporting an opinion and explaining the uncertainty within individual measurements.
Bringing the Evidence Together
The EDR steering and speed data alone did not show a clear path for the northbound car without expert review. The answer came from looking at different possible steering paths, speed data, damage orientation, and roadway evidence together. All these sources pointed to a crash sequence very different from what police first found. The reconstruction matched all available streams of data, including the crash severity measured by the northbound EDR. A simulation of this crash is shown below using Virtual Crash.
Full view simulation.
Scroll to move forward or backward through the simulation. You can also drag the progress slider or play the video.

Scroll controls playback position, not real-time vehicle speed. North points 10 degrees counterclockwise from the right. Frames extracted at 15 frames per second.
Zoomed view simulation.
Scroll to move forward or backward through the simulation. You can also drag the progress slider or play the video.

Scroll controls playback position, not real-time vehicle speed. North points 10 degrees counterclockwise from the right. Frames extracted at 15 frames per second.