Accident Frequency Rates

I’m sure we’ve all heard the saying “If you aren’t measuring it, you aren’t managing it” – if we aren’t measuring it, how do we know whether we need to improve it or, more importantly, how effective any improvements have been – but what are we supposed to be measuring?

While there are many things that can be measured and improved, one of the key metrics is Accident Frequency Rate or AFR. In the world of fleet safety and risk management we do try to steer away from the word ‘Accident’ – although commonly used, the word accident implies that whatever happened couldn’t have been avoided, and good fleet operators will more commonly measure ‘Incident’ rates.

Accident may be fine when referring to no-fault collisions – your driver may well not have been able to foresee or do anything about another road user behaving badly however, any incident where you accept blame is, by definition, not an accident. Whether it was caused by a vehicle failure, reckless driving, or even a simple lapse in driver concentration, the incident may well have been foreseeable and possibly avoided by better vehicle maintenance or more attentive driving – both of which are only improved through better management.

However even measuring incident rates can be confusing as there are different ways of measuring this. Some employers measure collisions per driver, some measure collisions per vehicle and some measure collisions per million miles. These rates measure frequency but some also measure severity by looking at the cost of collisions per driver or million miles. Many of the larger employers with good management strategies, including our own Driving for Better Business champions, will measure all of those metrics and monitor them month by month, carefully tracking trends and acting quickly when they see incident rates going in the wrong direction.

Once you’ve started measuring incident rates, the picture probably doesn’t become any clearer. You’ve arrived at a figure of 1.27 or 5.3 but what does that mean? Is it good or bad? What are you supposed to do now? What happens next depends largely on what size of company you are.

Having calculated your incident rate, there are three things that you could look at next.

The first is to continue monitoring month on month, year on year, to detect any trends and to understand the reasons behind those trends. Failure to understand why trends are moving in a particular way could mean expensive and misdirected interventions with no noticeable benefit.

When looking month by month you might see seasonal variations depending on your business. These may be due to weather making driving more hazardous or it could be that your drivers are covering more miles during busy periods. This is why it can be beneficial to look at collisions per million miles rather than focusing on how many incidents a particular driver has had. A driver with two collisions in 100,000 miles isn’t twice as high risk a driver who has only had one collision but has only covered 10,000 miles.

Some of our Driving for Better Business Ambassadors are big advocates of benchmarking. Colin Knight, Head of Fleet Safety Management and Compliance at the Clancy Group, monitors incident rates on a monthly basis:

It’s also important to understand that variations can occur for other reasons such as structural changes in the business. Rory Morgan, Head of Logistics Support for Western Europe at data management firm Iron Mountain experienced one such blip.

Once you have your incident rates, you can also start to benchmark them against others. If your organisation is big enough, the easiest way to do this is between different depots or groups in the same organisation as you can control the calculations to ensure consistency across the group. Benchmarking incident rates within the same group really highlights where improvements can be made and is an excellent opportunity to create some competitive spirit to help raise standards. Iron Mountain’s Rory Morgan thinks this is a valuable exercise.

One of the big benefits of internal benchmarking between different sites or depots is the ability to highlight consistency issues. A site with a lower score could indicate one of a range of issues including poor people management, policy implementation, induction process, training, etc. Benchmarking might not highlight the exact reason for the difference in performance but is certainly points you in the right direction for further investigation.

Benchmarking within your own organisation offers some useful lessons however the holy grail is to benchmark your organisation against other similar ones but this is the most difficult to do. The problem is that so many employers measure and monitor their incidents in different ways so you could find yourself comparing apples with oranges. Added to this, most fleet operators are extremely reluctant to share information that may highlight their performance being below that of their peers.

One route you could take is the fleet benchmarking assessment at www.fleetsafety benchmarking.net created in a collaboration between eDriving, RoadSafe, RoSPA and Driving for Better Business.

This will show you where your incident rate sits against other organisations that have completed the survey. Using this survey, the average collisions per driver stands at 0.09, collisions per vehicle at 0.12 and the average figure for collisions per million miles stands at 3.77 however we would sound a note of caution in that these companies may have different transport characteristics to your own.

As we pointed out at the start, if you aren’t measuring, you aren’t managing. If your figures compare well with those above then you’re probably doing well. But if your incident rates are on the wrong side of those above, congratulations! You might not be quite there yet but you could well have just highlighted an opportunity for making some significant improvements to the way your fleet is run.