Solving Problems with Crash Detection Technologies

Editor's note: This is the first part of a two-part series.

Black box recorders have traditionally been associated with aeroplane accidents helping investigators to determine what went wrong. Similar technology is now extensively utilized by insurance companies and road accident investigators to ascertain what actually occurred in major car crashes. However such technology has not been able to accurately determine what has occurred in low level crashes. Now such incidents can be detected and analysed with new, advanced, crash detection technology.

When vehicles are involved in a 'serious' crash (defining serious as bent metal or personal injury) there is little doubt that the event has occurred as significant, easily measurable G-forces are generated by the impact. The vehicle(s) stop, names and addresses are exchanged and an insurance claim usually results.

However, this is the tip of a cost iceberg as far as insurers are concerned. Many accidents do not involve bent metal or personal injury (dents and scrapes also lead to costly claims). Also, insurers are exposed to unknown costs due delays in claims.

[Canada: The Next Telematics Insurance Battleground?]

Detecting low level crashes is difficult. Forces created by bad driving (for example, curbing a front wheel at the approach to a roundabout) or bad road surfaces (potholes) can create forces similar to a minor crash. Neither owners nor insurers wish to be troubled with false alerts to incidents. So the technologist is stuck with a problem - either setting the G-force thresholds to a low level, resulting in annoying false reports, or setting the limits higher, thereby missing some crashes. For example, a speed bump may generate a vertical acceleration of around 3g, a force that could correspond to an impact. Taken in isolation, this event can be eliminated as a result of having no associated acoustic component.

Advanced crash detection technology now exists that operates by using two independent methods for detecting impacts. The first approach detects G-forces acting on the vehicle in three axes x y and z (fore and aft, lateral and vertical respectively) using an accelerometer array. The second method detects acoustic waves transmitted through the body of the vehicle. A collision is indicated only when both specific thresholds of both the acoustic and g forces are met.

Conventional methods use only the accelerometer technology, which creates a problem in discriminating between environmental factor (such as speed bumps, potholing and kerbing) and low speed impacts. Many of the early attempts to detect crashes produced as many false reports as valid ones, creating difficulties for insurers and insured!

In summary, high-speed impacts involving high energy are easy to detect. Low speed impacts are hard to differentiate from bad driving. The ability to differentiate is a key factor for insurers.

Next time, we'll discuss how this technology works, how it's been tested and applied in the real world, and how it works in concert with other data to provide insurance companies a full view of an accident.

About the author: Wayne Gilbert is CTO of Risk Technology Ltd, a telematics product provider based in the UK.