Most police departments have one or more officers who specialize in the detailed investigation of traffic accidents.  Through these investigations they are able to reconstruct the events of the accident.  It is no small task and it carries with it an enormous burden to be thorough and accurate in your conclusions.  Nearly all accidents deemed serious enough to warrant this kind of detailed investigation are destined to end up in criminal court, or civil court, or both.  And in each court there's going to be at least one expert witness on the other side, looking to pick apart your investigation.

In Brookfield I am in charge of the Accident Investigation Team, and I am also one of two reconstructionists on the team.  We are on call 24/7 and we respond to all motor vehicle accidents in town in which there is a loss of life, or where the injuries are so severe that the party may later expire from those injuries.

The job of a reconstructionist is in three parts.  First he or she has to respond to the scene of the accident.  Once there they must collect evidence as well as locate and interview witnesses.  Measurements are taken of the roadway and the vehicles and objects involved, head lamps and taillights are examined, crush damage to vehicles is profiled, vehicles are inspected and analyzed, the slope and drag factor of the roadway is measured, and everything is photographed and videotaped.  Normally it takes our team of five people between three and four hours to process the scene of a fatal accident.  There's a lot to do and you can't afford to miss anything.

Next, the reconstructionist has to prepare a detailed report of what occurred in the crash.  This includes a scale diagram and may include a vector diagram as well.  All pertinent information must be included in the report because it may be years before the case reaches civil court.  You can't count on memory alone to remember the specifics of a crash from ten years ago.  The movements of each vehicle and pedestrian before the crash, during the crash, and after the crash must be plotted with precision.  Time-distance analysis of all involved traffic units must be completed along with a number of "what if" scenarios, such as:  "What if Vehicle #1 had been traveling at the speed limit of 30 MPH instead of at 47 MPH?  Would this accident still have occurred?"

Finally, the reconstructionist must be prepared to testify in court.  Because of the training the reconstructionist has received he or she may be offered as an expert witness.  An expert witness is much different than a regular witness in court.  A regular witness may only testify to direct evidence, that is, something that they themselves have seen, heard, or done.  They cannot speculate or offer their opinion.  An expert witness can.  An accident reconstructionist is often required to analyze the facts and circumstances in the crash and offer their opinion of what happened.

In Connecticut this sort of work is extremely important to the insurance industry.  Connecticut resolves civil liability cases in percentages.  This means that if in a particular accident one party is suing the other for a million dollars, and the defendant can show that the accident was only 75% his fault, then the defendant just saved himself (or, rather, his insurance company) a quarter of a million dollars.  Insurance companies regularly hire (or keep on retainer) accident reconstructionists to do exactly this sort of thing.  I'd love to get into this aspect of accident reconstruction work - it would be an interesting way to supplement my rather meager police salary.

The actual reconstruction of a traffic crash is somewhat easier that it sounds provided you are highly organized, detail-oriented, and have a good head for algebra, trigonometry, and physics.  All of the objects involved in the crash (passengers, cars, motorcycles, pedestrians, mirrors that break off, cargo that gets ejected, etc...) are going to adhere to Newton's three laws of motion.  Knowing this, I can recreate their paths with certainty.  All I need is a starting point of some kind, such as a slide to stop skid mark, a critical speed yaw mark, an airborne launch point, or a point of impact with discernible approach and departure angles, and I can recreate the whole accident.

Click here to take a look at some of the equations we use in traffic crash reconstruction.

One thing I've noticed since becoming an accident reconstructionist is that most, if not all, of the people involved in an accident do not give me an accurate recounting of what happened.  Sometimes this is due to the fact that the person has just been through a stressful, traumatic incident and they cannot clearly recall what happened.  But sometimes it's because they know they did something wrong, and they figure that since I wasn't there to see what happened then obviously the only way I can find out what happened is if they tell me.  They assume that I may be able to take an educated guess as to what happened, but I'll never be able to prove it unless they confess.  And since they know that police officers have to deal with what they can prove, and not what they can guess, they won't be found at fault for the accident as long as they don't admit to whatever it is they did wrong.  So they make up a story and stick to it no matter what.  It's the sort of thing that always brings a smile to the reconstructionist's face.

I always ask drivers and witnesses for their version of events, but I rarely give much weight to what they say.  The physical evidence at the scene, properly interpreted, will never lie.  People sometimes will, especially people who know they are at fault and who don't want their insurance rates to go up.

If I respond to the scene of an accident in which you drove off the road and hit a tree, and once there I measure 130 feet of skid marks left by your vehicle, it will make me chuckle to hear you solemnly swear that you were only going 25 MPH when a deer ran out in front of you, and you are positive that's how fast you were going because you just happened to look at the speedometer just before the deer jumped out of the woods.  Unfortunately for you I know and I can prove in court that on a road with a drag factor of 0.75 you were traveling at a minimum of 54.08 MPH when you started to skid.  And let's say that after hearing your ridiculous story about what happened I measure the crush damage to the front of your vehicle and, from that, determine that you were traveling exactly 15 MPH when you struck the tree.  From that starting point I can work backwards and determine that you were traveling at exactly 56.09 MPH at the start of your skid, and that exactly 2.49 seconds elapsed from the moment you started to skid until you slammed into the tree.

Not one small part of those conclusions is guesswork on my part.  I can prove it all mathematically in court, beyond a reasonable doubt.  None of your emphatic denials to the judge are going to change the laws of physics.  People tell me all the time that I wasn't there and didn't see the accident and therefore I can't possibly know what happened.  It just isn't so.

I do occasionally make educated guesses at accident scenes, but those guesses are always based on my training and experience.  At less serious accidents, such as those where no one is injured and where there's only minor damage to the vehicles, it simply isn't feasible to close the road for hours and call in a half a dozen guys on overtime while we take the measurements necessary to mathematically reconstruct the entire accident.  In those instances I can look over the scene and figure out what happened even without taking precise measurements and doing the math.  I wouldn't make guesses as to exact speeds or things like that, but I can certainly fill in the broad strokes of an accident.  In the example above, with the car into the tree, I would know just by looking at the skid marks that the driver was lying about how fast he was going.  Without taking measurements I would still be able to guess that the driver was traveling more like 50 or 60 MPH, and I'd be right.  There's nothing mysterious about it.  I have been to advanced classes and I practice my formulas constantly using case studies in technical journals; I also practice my at-scene skills nearly every day in the field.  With that kind of background it would be odd if I couldn't make accurate guesses as to what happened after examining the scene of an accident.