Tools for Accident Reconstruction

The building of an accurate working model starts with a comprehensive measurement of the scene, as it currently exists. Recent developments in 3D laser scanning have produced instruments that can scan 1 million measurements of a scene every 10 minutes, within 6 millimeters of accuracy. The laser can scan a scene from a distance of 500 feet, while cars and people are moving through it. The density and accuracy of the measurements provide a realistic 3D image of whatever the scanner sees, allowing the forensic animator to develop an accurate computer-generated 3D model, without closing the scene to traffic.

The working model can be modified using photos and laser-assisted photogrammetry to match the scene as it existed when the event occurred. For example, if a building in the original scene has been demolished and replaced with a new one, photos of the original building could be used to place it in the working model of the scene. Using laser-assisted photogrammetry, the dimensions of the recreated building would not only be accurate, its position in the model would be precisely as it was during the event.

Physical evidence removed from the scene also can be reconstructed from photographs. Not only can the evidence be accurately modeled, but it can be placed into the working model in the same place that it occupied when the event occurred.

In a recent case, a photo of a skid mark taken at the scene of an accident that occurred three years ago was used to recreate the skid mark and place it in the working model of the accident scene, even though the mark had been erased from the intersection years earlier.

And … Action

Once all stationary elements are placed in the model, the reconstructionist can work with the forensic animator to ensure that all motion, including velocity, acceleration, and vectors, are added. Off-site tests may be used to determine some of these variables.

Once the model has been reviewed and validated by the expert, it can be used to simulate many different scenarios of an event. Using different variations of critical variables, including speed, acceleration, and trajectory, the expert can determine which scenario most closely correlates with the physical evidence, such as dents, paint transfers, and skid marks. Conversely, the expert also can determine which scenarios are not supported by the physical evidence.

The 3D model allows the expert animator and the attorney to see the event as it occurred and to understand how all the components interacted. In this way, the reconstruction team can ensure that all the factors are consistent with proven science and mathematics.

The location of each eyewitness to the event also can be placed in the model. The expert then can use the working model to determine what was available to be seen and what could not have been seen, for each witness.

All these analyses can be done on a computer, which substantially reduces the cost and time required, compared with on-site simulations. An additional advantage of computer-generated 3D reconstruction is that, when new data become available, the model can be updated to include that information. Additional scenarios then can be developed and tested, quickly and inexpensively. When the opposing attorney develops new ideas of what could have occurred during an event, those hypotheses can be tested in the model to determine whether they are possible.

The computer-generated model can be used to determine bullet trajectories and shooter locations from bullet casings and bullet holes. In a recent case, police used a working model to determine that a bystander had been shot by a ricocheting bullet from an officer's revolver.

Case History

The use of laser scanning and laser-assisted photogrammetry to develop 3D models of events changes the way that reconstruction experts and animators work together. A much closer working relationship between the experts and the animator is necessary to ensure the accuracy and the admissibility of the resulting testimony and animation. The animator must understand the science and mathematics that the expert uses, and the expert should understand the strength and limitations of laser-assisted photogrammetry and computer-generated 3D animation.

Experts and animators are continuing to develop new ways to use forensic animation to accurately reconstruct events. In the following example, the defense successfully settled a construction accident case using 3D modeling techniques.

A construction worker was preparing a large rebar cage to be lifted into place. Behind the worker, a front-loader operator was moving a large pile of aggregate. The operator backed up the front loader, crushing the construction worker. Witnesses to the accident offered conflicting testimony.

The attorney for the defense needed to know whether the operator of the front loader had seen the construction worker as he backed up, and whether the construction worker could see or hear the front loader backing up. Other questions to be determined included who was at fault and whether shared responsibility could be determined.

To determine what was available to be seen by the operator and the injured worker, the construction site had to be reconstructed, as it existed when the accident occurred. The dimensions and location of the rebar cage also had to be determined with sufficient foundation to be admissible.

The construction site was scanned using a laser to develop an accurate working model, including adjacent houses. An aerial photograph was obtained, which included the construction site as it was four months before the accident occurred. Using photogrammetry and other software, the dimensions and location of the rebar cage and the gravel pit were calculated and placed into the working model.

The front loader also was scanned, creating a 3D model. The front-loader operator was scanned sitting in the front loader, which provided the correct eye position for the working model.

The speed, acceleration, and path of the front loader, as determined by the off-site reconstruction, also were placed into the model. The animator and expert jointly developed a human head-and-eye jig and applied the head-turning movements as described in the testimony.

The expert calculated that the operator had a 160-degree field of view, which was used to determine the operator's view cone. The result was a time-synchronized reconstruction of what was available for the operator to see.

The final information needed was the location of the construction worker when the front loader backed up. Based on eyewitness testimony and the off-site reconstruction, the expert determined the construction worker's path and speed (4.4 feet per second).

The model now could be used to simulate what the front-loader operator actually could see as he turned his head left and right while backing up. The animation clearly showed that the operator could not see the construction worker at any time during the movement of the front loader. The construction worker could see and hear the front loader.

The combination of the blocked view of the operator and the worker's failure to react to the machine's size and noise as it backed toward him, provided a compelling reason to assign comparative negligence to the plaintiffs.

In mediation, the plaintiffs, who never admitted to comparative negligence by way of Requests of Admissions, lowered their demand more than one-third of the pre-mediation demand. In this case, the defendants saved in excess of $2.3 million dollars.

The working model provided the defense attorney a defensible position and two strong animations of the accident, for a cost substantially less than an actual simulation on-site, even if such a simulation would have been possible. s

Craig Fries is president and CEO of Precision Simulation in Grass Valley, Calif. He can be reached at craig@precisionsim.com.