A few days ago, six of us had a free ranging Zoom talk about using animation to determine the cause of a traffic accident. It was prompted by questions like the following:
- Is there a problem using animation in forensic work?
- What’s the problem?
- How serious is it?
- What’s the cause of the problem?
- How can the problem be fixed?
I learned from the talk that in spite of the mess at the scene of a traffic accident there are reliable methods that experts use to determine the cause. These methods involve modeling and animation software that are input with data and measurements got at the scene.
But, misleading data can also get into the software. It was this fact that prompted one person to suggest blogging about animation in forensic work. (Refs 1, 2) Another suggested a Zoom talk about it. It’s a problem because once the participants in the judicial and dispute resolution processes see a misleading animation they can’t un-see it – the impression has been made.
Participants who might be misled include:
- Insurance claims adjusters
- Civil litigation lawyers
- Judges and jurors
- People involved in alternate dispute resolution
- Injured parties in general
The data obtained during an accident investigation is put into modelling or simulation software that develops an inanimate model picture of the accident scene that includes the vehicles and features involved. Some of the data also goes into animation software connecting this software to the inanimate model making the vehicles appear to move.
Examples of animation:
The cartoon figures you see in TV advertisements started out as inanimate pictures in modelling software. Animation software got the figures moving to amuse and catch our attention – and prompt us to read the advertisement.
The same principles underlie the animation of a traffic accident that helps us understand what happened. Some animations are alarming to watch and not amusing.
You might take a look at the following simple animation depicting an accident from different view points and by the different drivers involved in the accident. It will give you an idea of what I’m talking about. If you are somewhat in awe at what you see then you understand part of the problem with animation in forensic work:
Talkers at the Zoom table:
- Two traffic accident investigators (Reconstructionists; the name in common use in North America),
- A former accident investigator who now lectures,
- A fourth chap who also investigates failure in general as well as traffic accidents,
- A fellow who animates the results of traffic accident investigations,
- And your’s truly who investigates the cause of failure in the built and natural environments.
Stages followed during an accident investigator:
- Collect data. Data like published mapping and photography of the site including a Google Earth image and/or drone photography. Walk and visually examine the site. Measure and photograph features at the site, and, Read police reports on the accident.
- Analyse data. Draw diagrams of the accident site. Place the objects (vehicles) and features associated with an accident on a Google Earth image to produce a land surveyor-type site plan for use in the simulation/modelling software. Model the traffic accident using published software. Animate the model using software to better understand the data and to ‘give it life’. Add colour, lighting, texture, etc. to the animation software. Study and note the different causes of the accident as indicated by the individual data. Note the extent to which one cause of the accident is indicated by the bulk of the data.
- Draw conclusions. Conclude the different causes of the accident as evident in the data and why one cause is more probable than another.
- Form opinion. Draft opinion on the cause of the accident.
- Write forensic report Write report in technical English, plain English or a combination technical/plain according to the requirements of the client.
Some questions asked by a traffic accident investigator:
- How many vehicles were involved?
- What kind of vehicles were there?
- Which vehicle crossed into the other lane?
- How did they get the exact positions of all vehicles?
- How many lanes were?
- How big was the shoulder?
- What were the lane markings like?
- Were they worn?
- Were there tire skid marks and where were they located?
- How long were the marks?
- What was the skid resistance along the marks (coefficient of friction in high school physics)?
- Were there pedestrians or other obstructions on the road?
- What was the weather like (clear, rain, snow)?
- What time of time was it (night, dawn, day)?
- What was the visibility like?
You can easily imagine how careful a reconstructionist must be with so much data to collect and input. S/he must know how to use the software correctly, know the correct values to input, and how to correctly interpret the results. There’s no room for error in a forensic investigation when giving an opinion on cause.
- Soft data like visual effects (lighting and colouring) can also be useful if added to the software. But, you’ve got to be careful with this one.
One talker at the Zoom table noted: “I think whether it’s animation or simulation/modelling, the input values need to be shown to be correct, or at least reasonable, for the results to be correct and admissible. My concern is when realistic but unproven graphics such as lighting and colouring are used. These effects can leave a subconscious impression on a judge or jury that can affect their perspective and judgement.”
Accurate animation software
Most accident reconstructionists accept that the modelling and animation software programs most frequently used are accurate. These are PC-Crash and Faro Blitz to name two of three or four. Nevertheless, where possible, calculations are checked by hand and input values made certain.
Misusing animation software
It’s also at the modelling and animating stages that inadvertent or deliberate misuse can occur. The talkers at the Zoom table didn’t think the latter was a serious problem.
Still, the software sellers “build a case” promise still rings in my ear, and how that must resonate in the ear of some bad-cat experts. Google forensic animation and see for yourself. I am comforted by Ruth Corbin’s findings in her excellent pilot study of 152 experts that all experts clearly understand that they serve the judicial process not their client or themselves. (Ref. 3)
Expert reporting in plain, jargon-free English
With so much data/input arising from a traffic accident reconstruction I believe expert reports in plain English are essential for non-technical readers, or at least a summary report. Animations are so realistic – and in many ways, entertaining like a roller coaster – it’s essential to help readers know what they’re seeing and must focus on.
What did I get from our talk?
- Animation in forensic work is helpful in resolving disputes, particularly in determining the cause of traffic accidents, but keep it simple.
- Don’t submit the animation as evidence unless it’s a simple animation.
- Check the input data. Check the output animation when possible.
- It’s difficult to correct a false impression taken from an incorrect animation.
- Modelling and animation software are quite accurate.
- But remember that the software seller is in business to sell software, and help you “build a case”, as s/he sees it.
- Animation in Forensic Work: Use and Misuse. Posted January 12, 2022. Updated January 20, 2022
- Telling it like it is, in forensic expert report writing. Posted January 27, 2022
- Corbin, Ruth M., Chair, Corbin Partners Inc. and Adjunct Professor, Osgoode Hall School, Toronto, Breaking the Expert Evidence Logjam: Experts Weigh In, presented at Expert Witness Forum East, Toronto, February, 2018 (Google this excellent paper and Ruth’s name)
(Posted by Eric E. Jorden, M.Sc., P.Eng. Consulting Professional Engineer, Forensic Engineer, Geotechnology Ltd., Halifax, Nova Scotia, Canada, February 27, 2022. firstname.lastname@example.org)