Setting the Scene – Using Drones for Traffic Accident Reconstruction and Analysis
Greg Gravesen is an experienced traffic accident reconstruction specialist. Here he explains how employing a fixed-wing mapping drone has revolutionised his work.
Hi Greg, why don’t you start by telling us a little about your background and experience?
Hi. I’ve been a law enforcement officer for 25 years. During the past 22 years, I have developed a specialty in the area of crash reconstruction. In addition, through my position as a Sergeant with the St. Paul Police Department’s Forensic Services Unit, I have had extensive experience in the area of crime scene documentation and illustrations.
How has the technology you use in this work changed over that time?
Early on in my career, I used tape measures to document my crash scenes. As technology developed and the costs got lower, I eventually upgraded to a laser scanning system. The past 15 years, I have incorporated terrestrial surveying instruments such as a Trimble robotic total station and a FARO scanner to document both crash and crime scenes. I am also involved in teaching this technology. Earlier this past year, I purchased the senseFly eBee drone, which has become my primary methodology of scene documentation. This system allows me to collect information for very large scenes very quickly.
How do you use the geospatial data you collect?
The senseFly mapping system allows me to create the orthomosaic image, a 3D point cloud, and a 3D mesh of my crash and outdoor crime scenes. This output can be used in a number of ways, from a pure graphical perspective and also during the analysis.
The eBee produces a high-resolution image that normally does not contain moving vehicles. In other words, if a moving vehicle is contained in one image, that cell of the mosaic can be replaced by another image. This results in a stunning image that can be a trial exhibit. This image can be used in several ways. One of the most common applications is to use this image to create highly detailed and accurate scale diagrams. In addition, the orthomosaic may also serve as the background for two-dimensional simulations and animations. The resulting graphic is very visually appealing.
“This screenshot is an example of how you can bring photo realistic models, like cars, trucks and pedestrians, into a point cloud to create very realistic and accurate forensic animations. I use 3D Studio Max, a high-end 3D animation and illustration program that gives me complete control over movements, cameras, materials, textures etc.,” says Greg.
From a graphical perspective, I create a number of animations using the 3D point cloud data that I collect, for example with the laser scanner or more recently with the eBee. The most common type of animations involves a virtual “fly-through” of the scene. This allows the viewer to get a unique perspective of a scene that simply can’t be accomplished through traditional technologies such as photography. Using the eBee system these “fly-throughs” are easy and quick to create. In addition, I create more photorealistic animations where I insert scaled models of objects, pedestrians, etc. into the 3D point cloud. The 3D point clouds can also be used as a “blueprint” to create virtual models and environments using software such as Autodesk’s 3D Studio Max.
The video is a point cloud fly-through animation that depicts a scene over a mile long, which involved two vehicles drag racing. This animation allows the viewer to get a unique perspective of the entire crash scene. It would have taken days to use a terrestrial scanner to complete this project. This type of exhibit is really effective evidence. It gives juries and clients a unique perspective that traditional methods, such as photography, just cannot compete with.
It gives juries and clients a unique perspective that traditional methods, such as photography, just cannot compete with.
Another application of the drone’s output involves the use of computer simulation software, a powerful tool used in traffic accident reconstruction. This software predicts things such as vehicle positions, trajectories, acceleration, collision forces etc., which are based on Newton’s Laws of Motion. Specifically, I use Engineering Dynamic Corporation’s HVE and PC-Crash software packages. Both packages require an environment be created and later displayed in the simulated results. The orthomosaic image and the 3D point cloud created by the eBee are easily imported into this software and help produce a high quality, accurate output.
For example, the simulation video shows full and half-speed simulations I produced using HVE-2D software. Here I was able to bring in a drone-produced GeoTIFF orthophoto and simulate the collision over the top of it. If you look closely, the vehicles’ movements, post-impact, correspond to the actual scene evidence.
Why did you choose the eBee drone specifically? What made this system stand out?
I researched the eBee and was really impressed with it. However, I never really saw its full potential until I went out and bought one. It is an extremely versatile system that produces accurate results, quickly. Whether I am using the high-resolution orthomosaic image to create diagrams, the 3D point cloud to create a simulation environment, or creating fly-though animations, this work is done much quicker and safer than using other methodologies. For example, for large scenes when I would previously have used terrestrial scanners these would have taken one to two days to document (not including computer processing time). These take literally less than 30 minutes to fly and about three hours to process on the computer. Initially, the drone simply sounded too good to be true. However, I was amazed when I experienced it first-hand.
One of the other things I really like about the eBee system is its reliability and built-in safeguards.
What do you feel the acceptance of drone technology has been like in the crash and crime scene analysis world up to now?
Interestingly, rotary systems are becoming common, but some people think that fixed wing UAS have no place in forensics. I disagree completely. Often times, especially from the standpoint as a private consultant, my involvement is long after the date of the crash and I don’t have the luxury of shutting down a road. A product like the eBee excels here because I am able to capture highly accurate information about the scene both quickly and safely. I am able to obtain 3D data in a fraction of the time it takes with other terrestrial surveying instruments. And best of all, I do not have to enter the roadway or shoulders to obtain it.
… some people think that fixed wing UAS have no place in forensics. I disagree completely.
The resolution of the eBee’s imagery means I can actually identify the paint marks that law enforcement made on the ground, tire evidence, ruts in ditches, and other important evidence. Another strength of the eBee system is the ability to document very large scenes. For example, I recently documented a crash scene that resulted from two vehicles racing. The incident occurred over a one mile section of roadway. I was able to document this scene with one 24-minute flight. I am not aware of a rotary system that could cover that extensive a scene in one flight. The multi-rotor systems I am familiar with do not have the battery life that permits them to have the kind of range seen with the eBee system.
You’re working on a comprehensive study about your use of drone technology in crash reconstruction. Could you tell us a little more about what that will contain?
That’s right. I have already done some validation testing prior to using the eBee system for case work. I am now planning on conducting a more extensive study that will scientifically evaluate drone technology in the content of scene analysis work.
This testing will involve evaluating and comparing technologies such as the eBee, total stations and terrestrial laser scanners, and using a tape measure, in terms of their accuracy and efficiency. This type of research is useful to validate tools like the eBee and defend them with Daubert challenges in court. We hope to publish these results in a peer-reviewed journal.
Based on my experience so far, the drone will prove to be the fastest, most comprehensive data gathering tool. Its speed and accuracy, from what I’ve seen so far, are just incredible.
Greg, thanks so much for this overview. Do let us know when this study is complete.
You’re welcome, I will do.
Greg Gravesen has been an ACTAR-accredited traffic accident reconstruction specialist since 1994. He is the president of the private consulting company Northwest Crash Analysis, LLC and currently works as a Sergeant in the St. Paul Police Department, Minnesota, assigned to the Forensic Services Unit. He is also an adjunct instructor with the University of North Florida – Institute of Police Technology & Management.
Latest blog posts
Drone Mapping After the Storm: USAA Delivers Localized Disaster Intelligence and Claims Support with eBee X
The 2020 Atlantic hurricane season was one of the most devastating and costly on record. When Hurricane Delta made its arrival on the southern shores of the U.S., it broke the 104-year-old record for the most named storms in a season. It brought with it wind speeds of 145...
eBee Drones Help GeoAcuity Rapidly Survey 20,000 Acre Wind Farm Site
When CSRS needed to pre-survey a 20,000-acre site for a future wind farm in Tunica Mississippi, they turned to veteran-owned consultancy GeoAcuity for its 20-years of specialized geospatial experience and delivery of accurate aerial surveys under tight timeframes. Looking out your window traveling on U.S. Route 61, you can...
Talking Drone Training & senseFly’s e-Learning Platform with Andrea Blindenbacher
With the launch of senseFly’s new e-learning platform and dedicated Certified senseFly Operator Program, Waypoint recently sat down with senseFly Global Head of Training Andrea Blindenbacher to learn more about how the new platform and certification course works, where to access it and how senseFly users (and even non-users) can benefit from the various self-guided tutorials...