An example of a point cloud of a collision scene. Parasite images of troopers are caused as they move through the scene during scanning.

Like many law enforcement agencies, Washington State Patrol (WSP) has employed laser scanning as a tool for scene measurement and documentation. Rod Green, a 22-year veteran of WSP, has conducted accident and crime scene reconstructions for more than 14 years. In addition to collecting and analyzing information, he works to improve the reconstruction process on-scene and in the office. He has taken the use of scanning to new levels, and the results of his efforts have drawn attention in Washington and surrounding states.

Green’s work is unique because he combines experience and knowledge to simplify and expedite the process of collecting and analyzing information and evidence at accidents and crime scenes. He uses high-tech hardware and software tools with logic to make major improvements in reconstructing the events. 

Green has worked in every area of reconstruction, from data collection to the production of drawings. But, the technology doesn’t change the core mission of public safety and law enforcement.

“I’m not so concerned with how pretty the point cloud looks as I am with getting the details—all of the details—to accurately gather measurements of the scene for reconstruction,” Green said. 

Green believes that laser scanners’ abilities to gather huge amounts of information can play an important part in producing reports, drawings and exhibits that contribute to the successful resolution or prosecution of a case. But for scanners to become common tools, they must easily blend into the work of people who are already very busy.

Speed and accuracy

Investigators onsite of fatal collisions and crime scenes face several challenges. They need to collect comprehensive information in an accurate, systematic manner. Often working at night or in inclement weather, they must capture data quickly while the scene is fresh and before rain or snow can destroy evidence. 

Quick work is also critical to reducing the length of road or lane closures, which are costly and inconvenient to the public. According to a 2015 report by the National Highway Transportation Safety Administration (NHTSA), costs attributed to congestion caused by fatal crashes in 2010 including travel delay, excess fuel consumption, greenhouse gases and pollutants totaled $189 million. On average, a fatal collision resulted in a closure of nearly 3.75 hours. By using laser scanners for data collection, Green and his WSP colleagues showed that they could reduce the time on site while increasing the quality and quantity of information they gathered. 

In addition to on-scene measurements, reconstructors must manage and process the data to produce the information needed by investigators and district attorneys. They need to focus on the pertinent details while maintaining data integrity and the chain of evidence. Their finished products are exhibits, reports and analyses that provide the basis for the accurate depiction of the scene and events. Green has developed a combination of hardware, field procedures, software and data management that reduces time on site and produces finer detail throughout an investigation.  

The importance of scanning 

Scanners have been used in scene reconstruction for nearly 20 years. A scanner uses laser technology to collect a large number of closely spaced measurements and produce a “point cloud” of 3-D points. By setting the scanner in several locations around a scene, investigators can capture multiple point clouds to develop a 3-D snapshot of the location. Even small or subtle features such as shell casings or tire marks can be captured using the scanner and point cloud.

The scan data is then transferred to computers for archival and processing. The multiple scans are combined into a single point cloud, which provides a complete view of the scene.

However, large point clouds come with challenges.

“There are people who look at a point cloud and say they can’t see the details,” Green said. “That is usually because it wasn’t scanned or processed correctly. We want to avoid the trap of not seeing the evidence for the cloud. We do not want just a huge cloud; we want better management of the cloud.”

The solution comes from integrating the processes, from collection through final reports. 

That integration begins in the field, where WSP uses a Trimble TX5 scanner. The scanner simplifies the work for scene investigators by providing preset scanning modes that capture data at different levels of resolution. Green has developed guidelines for WSP investigators that recommend the appropriate mode for capturing different types of details and evidence.

“We like the presets,” he said, “because we can teach our people what mode to use in a given situation without having to calculate resolution. [For example,] I might need to capture shell casings or faint tire marks; for that we might scan at high resolutions. But for other setups we can run at lower resolution, which is much faster.”

Data from the scanner goes to the office, where detectives use Trimble RealWorks software to combine the scans and begin analyzing the scenes. The software can use “plane-based” and “cloud to cloud“ registration to merge multiple scans—the technique helps reduce the need for on-scene investigators to handle the spherical targets commonly used in many scanning projects. Although target spheres are needed in some situations, Green said the plane-based and cloud to cloud approach is much simpler and faster for field investigators—so much so, it’s become the default method. 

Eliminating the targets is just one way the software component of the integrated system helps make the job of on-site personnel and office faster and easier. For example, even when working at low resolution the scanner may collect thousands of points on a road—far more than needed to correctly depict the surface. The software can reduce the density of the points to preserve the accuracy of the information while eliminating clutter that could potentially distract or confuse a jury. 

As soon as the main point cloud is in place, the office starts creating reports and diagrams. Using reconstruction software, Green can draft maps and models in 3-D. Points that are superfluous or irrelevant to the case can be moved to separate layers that are switched off, but are still preserved to meet the essential chain of evidence requirement and maintain integrity of the scene’s data. When the point clouds are completed, the clouds and analyses can be supplied to prosecution teams.

Investigators can add evidence markers, vehicle models, trajectories and more—the software can even simulate how cars are crushed in a collision. The information can be used to produce diagrams and animations for courtroom presentations or can be sent for specialized analysis. The result is a concise representation of the scene and events—all backed by detailed, defensible data.

One of scanning’s most valuable features is the ability to preserve a scene while it is fresh so officers can “virtually” revisit it after the fact. For example, an object not deemed pertinent early in an investigation may turn out to be important. By using the point cloud, investigators can return to the virtual scene at any time. They can use the virtual return visits to extract or confirm details on the object and its location long after the on-scene events and investigation.

What’s key to Green’s success is his understanding of what goes on at the scene and beyond. By using a holistic approach to reconstruction data collection and analysis, he is helping to simplify the work of officers in the field as well as in the office. The result is faster, safer work with better results.

Chad McFadden is the Forensics Business Area Manager for the Geospatial Division of Trimble, Inc. He has over 20 years’ experience in Land Surveying and GIS projects, including field, office and business development.