Monthly Archives: January 2018

Figuring out the cause of forensic messes is not easy and takes time

Posted on by

So much of what we do in forensic investigation doesn’t lend itself to clean and tidy text-book investigation and analysis.  It’s messy and difficult to wrestle to the ground as to cause.  It also takes time.

I thought this after I posted a blog two weeks ago on getting hard evidence from soft data; getting the speed of a vehicle in an accident from the images on mobile phones. (Ref. 1)  I blogged at that time on a lecture by Major Adam Cybanski, Ottawa, on a new forensic method for assessing the speed of vehicles in an accident. (Ref. 2)

If you could see some of Adam’s lecture photographs and video showing aircraft, cars and people flying through the air and crashing, you’re certain to wonder: How can you measure and analyse something like that?  Disturbing images.  Surely traumatizing for the forensic investigator to see on site.

There was one picture of an airliner – not the TransAsia flight 235 crash mentioned in the lecture but that one too – nearly vertical in the air, nose down and a few metres from hitting the ground.  An aircraft filled with passengers.

If you could see the condition of some cars after a traffic accident that Dr. Stuart Smith and others in CATAIR investigate, that includes doing crush measurements, you’re certain to wonder about this too. (Ref. 3)  I’ve helped Stu a couple of times do these measurements; not a pretty sight.

Less visually disturbing – unless you are the owner – seeing undulating floor surfaces in multistory buildings or bent steel beams in half-built bridges you might wonder: How can a forensic expert analyse odd failures like these?

I sometimes wonder too and I come from a civil engineering background specializing in geotechnical and foundation work – measuring and engineering the messy ground.  We get it done because we have our models and semi-empirical relationships developed over decades of field observation, research and engineering practice.

Models help us figure out the cause of messy failures and accidents but their development takes time and lots of thought and hypothesizing.  I’m not surprised it took Adam months to get his analysis time down to a few days from a few months. (Ref. 1) In fact, I’m surprised it came together so quickly for him.

These models and analytical procedures show up in textbooks but not right away, not until someone has figured out the mess from a lot of failures.  I had one client surprised that forensic engineers must sometimes research topics for which little is known or vary geographically.  He thought we knew everything – a nice image.

***

(CATAIR; the Canadian Association of Technical Accident Investigators and Reconstructionists)

(Model: A set of ideas and numbers, based on existing data, that describe the past, present or future state of something.  For example, how the economy might react to a change in interest rates or a type of failure or accident occurred in the built environment.  Models are updated with new data as required.  After Merriam-Webster dictionary, January, 2018)

References

  1. Getting hard evidence from soft data.  Posted January 10, 2018
  2. Cybanski, Major Adam R., Gyro Flight and Safety Analysis, Ottawa, 2017 http://www.gyrosafety.com
  3. Smith, Dr. Stuart, C. R. Tyner and Associates, Dartmouth, Nova Scotia  crtynerassociate@eastlink.ca

How do you get hard evidence from soft data?

Posted on by

How do you get evidence about the speed of a car from cracks in the pavement or trees at the corner?  How do you get reliable, quantitative evidence from the qualitative data on a mobile phone?  Evidence that will stand up in court?

This is being done now by a new forensic investigation method.  It requires knowing what to look for in the phone’s video, some measurements with a carpenter’s tape, Google earth, a little junior high math and lots of software.

It’s called video velocity analysis, a scary title but remember: Carpenter’s tapes, junior high math and modern technology make it happen.

Video of aircraft accidents taken by witnesses has been analysed since 2008.  The crash of TransAsia flight 235 on February 4, 2015 in the Keelung River shortly after takeoff from the Taipei Songshan Airport was caught on three separate witness cameras, and was subsequently analysed.  Video of traffic accidents is now being done the same way.

Reliable forensic evidence

The qualitative data on traffic cameras and car dashboard cameras is also being used, as well as mobile phones to learn the speed of a car at the time of an accident.  In fact, it was the cross checking of an assessment of car speed from three different sources of soft data – mobile phone, traffic camera and dash camera – in field trials with a car with a speedometer and GPS, that has demonstrated the accuracy of this new method.

How accurate is an assessment of car speed using these types of simple cameras?  Depending on the circumstances, within about 2 km/hour – pretty hard evidence in a court case.

Google Earth sometimes figures in this type of assessment as well.  How accurate is the quantitative data from the eye in the sky, kilometres high?  How about within a few centimetres on the ground in urban areas.

Evidence based on junior high math 

The basic principle is simple enough.  Measure the distance between two points on the ground seen in mobile phone video, note the time on the video for the car to travel this distance, divide one by the other – junior high math – and you’ve got your car speed.  Photogrammetry, the science of making reliable measurements using photographs, is sometimes used in this work but the principle is simple.

What does the analyst look for in the mobile phone video, traffic camera or dash camera?  Basically, anything on or near the ground that can give distance that can be correlated with time which is also taken off the camera.  Things like the distance between construction cracks in the pavement, dashes on road centre lines and lane markings. The analyst is also interested in anything that can be seen on Google Earth.

Car speed in accidents has been measured other ways for years and continues to be.  Using mobile phones is new, can be more accurate and provides an opportunity for the cross checking essential to good engineering and applied science.  I had the cross checking of data drilled into me by Major James A. H. Church when I studied land surveying at the College of Geographic Sciences. (Ref. 1)

Explained in Moncton, NB in 2017

This all came out in a lecture I took last fall in Moncton by Major Adam R. Cybanski, Gyro Flight and Safety Analysis, Ottawa, on a new technique for learning car speed in an accident. (Ref. 2)

Major Cybanski has been instrumental in developing the technique using the simple, inexpensive mobile cameras that are everywhere today.  He specializes in video analysis and accident reconstruction for aircraft and auto accidents.  Adam flew all manner of aircraft in the air force over the years, slowly moved into the investigation of the cause of aircraft accidents and more recently into auto accidents.

And used in court cases

He believes he is one of a few in the world using this analytical technique.  In 2016 he had one case, but this grew to 14 cases from different accident sites in the world in 2017.  Video speed analysis has been used in several court cases and as of March, 2017 has not been contested.

An analysis took him many months initially when he was developing the technique using witness video.  It is something he can now do in a few days, depending on the location of the accident and the features visible in the video.  Time is spent cross checking an analysis; getting the car’s speed from more than one camera.

Vehicle speed is just one element in accident reconstruction but an important one.  Adam gets camera data sent to him from the principle investigator wherever he or she is in the world, analyses the data and sends the auto speed back to the investigator.  He doesn’t need to visit the site in most cases.  His speeds have been validated using speed radar guns, GPS and Event Data Recorders – the “black boxs” installed in some cars to get information during accidents.

The lecture was organized by CATAIR, the Canadian Association of Technical Accident Investigators & Reconstructionists, Atlantic region.  Dr. Stuart Smith, secretary of CATAIR knew of Major Cybanski and his technique and suggested inviting him to Moncton to speak.  Stu reconstructs traffic accidents in his practice including analysing vehicle speed. (Ref. 3)

The take-away

What’s the take-away from this blog?  A new and accurate method is available to check vehicle speed in an accident as determined in more conventional ways.  The results are reliable and accepted as evidence in court.  And expertise is as near as your e-mail.  The speed of anything that moves can be analysed if caught on a mobile camera.

References

  1. Church, Major James A. H., founder and first principle of the Nova Scotia Land Survey Institute, 1947, forerunner of the College of Geographic Sciences, Lawrencetown, Nova Scotia
  2. Cybanski, Major Adam R., Gyro Flight and Safety Analysis, Ottawa, 2017 http://www.gyrosafety.com  (Adam and I conferred about this method as it’s new to me too)
  3. Smith, Dr. Stuart, C. R. Tyner and Associates, Dartmouth, Nova Scotia   crtynerassociate@eastlink.ca