Omnitracs' Road Ahead blog

Identify Crash-Prone Behaviors to Keep Fleet and Drivers Safe

Don Osterberg
Safety Advisor, Omnitracs

Regular readers of my blog posts are familiar with what I’ve called the “unacceptable increase” in truck-involved crashes resulting in injuries and fatalities.

From 2009 to 2019, the number of trucks involved in fatal crashes rose 56% (1), from just over 3,200 (2) to over 5,000. Fatalities rose 48% to 5,032 (3). The impact is financial as well. Insurance rates rose 62% from 2009 (4) to 2022, while the estimated cost of police-reported large-truck crashes was $91,112 in 2005 (5): adjusted to $140,000 today (5). In a recent blog about nuclear verdicts, I also mentioned the cost of so-called nuclear verdicts, with recent amounts climbing from $90 million to a billion dollars in just a few years.

The 2006 (6) large truck causation study by the FMCSA showed that in 88% of fatal and injury crashes, driver factors such as driving too fast for the conditions, illegal maneuvers, illegal drug use, lack of familiarity with the road, and driver fatigue all played a significant role.

Predicting Crashes by Driver Performance

A more recent study, Predicting Truck Crash Involvement (2022 Update) by ATRI (American Transportation Research Institute), looked at driver behavior as a predictor of crashes rather than a causation. If you could identify the drivers who exhibited the predictors, you could take pre-emptive action.

The report identified a database of 583,805 unique drivers: those who had undergone inspection during a specific three-month period (Jan, Feb, and March of 2019), as noted by MCMIS (Motor Carrier Management Information System). Information on inspection violations, either vehicle or moving violations, convictions, and crash history was gathered from both MCMIS and CDLIS (Commercial Driver’s License Information System) data. I know many of you would agree that MCMIS data is untimely, inaccurate, and inconsistent. I wouldn’t disagree with those who see it that way, but it’s consistently inconsistent and the best database available.

Of these 583,805 drivers, 38,797 were involved in crashes, about 6.6%, while the rest, over 93%, were crash-free. So let’s take a look at that 6.6%.

Of that 6.6%, 34,000 drivers had one crash; 1,881 had two, and only 81 drivers had three crashes. Six drivers had four crashes each, and last place was tied: two drivers with five crashes each. One can imagine learning a lot from just those last 89 drivers. But let’s take a look at the bigger picture.

What the Inspections Found

It’s always interesting to note what inspectors find during truck inspections, and here, the drivers performed better than the trucks. Violations included 291,000 for lighting, 250,000 for other vehicle defects, and 238,000 for brakes and other problems. Driver violations included Hours of Service (116,000) and speeding (51,000), plus size and weight, log book, and other issues. In all, the study incorporated more than 1.69 million violations among the 585,000 drivers.

Crash Predictors

The study found four specific violations, convictions, or historical occurrences that each raised the likelihood of a crash by over 100%:

  • Failure to yield right-of-way violation: 141%
  • Failure to signal or improper signal conviction: 116%
  • Past crash: 113%
  • Reckless driving violation: 104%

These convictions predicted an increased crash likelihood of 50% - 100%

  • Failure to obey a traffic sign
  • Failure to keep in the proper lane
  • Improper or frequent lane changes
  • Reckless or negligent driving
  • Traffic signal violations

Other significant predictors include speeding, HOS and log book violations, tailgating, and size/weight violations.

Intervening Before There’s a Problem

Interesting statistics. But, as I like to say, so what and then what? Now that we have statistics on crash predictors, what do we do with them? How do we use them preemptively to reduce the frequency of truck crashes and the injuries, deaths, and costs they incur?

Consider three steps:

  • Evaluate the drivers for these specific behaviors
  • Determine the root cause:
    • Did the driver not know these are improper actions and thus, a training issue?
    • Or did the driver know and proceed anyway, and thus a behavioral issue?
  • Choose the remedial path that best matches the situation

Evaluation: Video and Telematics

New technology has provided some excellent tools to objectively observe and evaluate fleet drivers with greater accuracy and detail than previously possible. Video technology, available inside the cab, road-facing, and multi-directional, can observe any number of risky driving actions, from distracted driving to tailgating, and then alert the driver and note the event. Fleet telematics registers driving patterns for signs of aggressive or reckless driving, such as speeding, hard braking, or fast accelerations. Either system is helpful; used in conjunction, the driver insights they provide can be extremely accurate and actionable.

Get to the Behavior’s Root Cause

You can’t fix a behavioral problem with training or a training problem with a behavioral approach. Successful interventions need to address the correct root cause, and here I find that the 80/20 rule generally prevails: 80% of crashes are due to behavior. About 20% are a skill or knowledge deficit. To get to the root cause, I often apply what I call the Five Why’s methodology. Here’s an example of the Five Why’s:

Problem: Driver Cited for Failure to Yield Right-of-Way

  • Question 1: Why did the driver fail to yield?
  • Answer: The driver was distracted.
  • Q2: Why was the driver distracted?
  • A: The driver was distracted by a cell phone text message sent by his/her dispatcher.
  • Q3: Why did the dispatcher text the driver while he/she was driving?
  • A: It was easier than checking the telematics to see if the driver was moving.
  • Q4: Why was it easier for the dispatcher to send a text message to the driver than to check to see if the driver was driving?
  • A: Because the dispatcher was task-saturated and was cutting corners by texting without first checking to see if the driver was driving. The dispatcher texted to check on the status of the load to ensure it wasn’t running late.
  • Q5: Why was the driver reading/responding to a text message on a cell phone while driving?
  • A: Because he/she had done it before without issue and was desensitized to the risk.

Root Causes:

  • Dispatcher taking the path of least resistance to save time and effort. [Behavioral]
  • The driver was desensitized to the risk and chose to violate company policy and FMCSA regulations by reading a text while driving. [Behavioral]

Interventions to Preclude a Recurrence:

  • Reinforce with the dispatcher that telematics, not the driver, is the first source of this and other data.
  • Reinforce with the dispatcher to NEVER text a driver while the vehicle is in motion, regardless of the reason (or if the vehicle is stopped and the driver is on the sleeper berth line).
  • Reinforce the expectation with the driver to NEVER read or respond to a cell phone text message while driving. Ask the driver to commit to changing his/her behavior. Have them sign a document committing to the behavioral change.

Decision or Awareness?

The texting was basically poor decision making all around. But even excellent drivers may have developed certain habits they are unaware of and even defensive about because they may not see things in context.

For example, I was involved in an accelerometer test, and we needed to calibrate truck standards for acceleration, hard braking, etc. I took our five best drivers, all million-milers, and put sensors on their trucks to monitor performance and establish the baseline. When it came to hard braking, one driver was noticeably over-represented, and yes, he was very defensive when I mentioned it. I responded by laying out the metrics of all five drivers, so he could objectively see what the difference was. He never said I was right, but he didn’t have to. By the next week, his performance was right in line with everyone else's because he had awareness and objective context.

Addressing Situations Preemptively

We seek out the root cause because root causes are actionable. You can effectively address predictors before they become incidents.

With the million-mile hard-braker, I didn’t lecture or reprimand or judge; I provided objective context. For the texting dispatcher and driver, it was a case of reinforcing expectations. And if it’s really a training or skills issue, then it’s a training solution.

Here is what I think is the best way to apply the Crash Predictors to reduce the likelihood of crashes:

  • Evaluate all drivers. It establishes a baseline and keeps any drivers from feeling singled out. And even your best drivers will benefit.
  • For drivers over-represented above the baseline, apply the Five Why’s and establish the root cause and address the actual situation.
  • Make sure to communicate to drivers that these measures are proactive and protective, not punitive. You want to keep them safe.

And do everything within a culture of safety.

“Nothing we do is worth harming ourselves or others” is an adage from early in my career that still holds sway. It says that safety is a core value, a part of the culture. Any efforts to improve safety will be much more successful within a culture that prioritizes safety at all points and all levels, including and especially from leadership. Within such a culture, all safety initiatives will have the greatest chance of success.




1. Traffic Safety Facts, April 2022 (2020 Data), Table 1, p 2; Table 2, p 3. NHTSA, DOT HS 813 286

2. Traffic Safety Facts, Large Truck Fact Sheet, Revised May 2014, 2012 Data

3. Deborah Lockridge, ATRI Takes on Rising Trucking Insurance Costs, Feb. 18, 2022 (Heavy Duty Trucking,

4. Unit Costs of Medium and Heavy Truck Crashes, (FMCSA, March 2007)

5. $92,000 in 2005 → 2022 | Inflation Calculator. (Official Inflation Data, Alioth Finance, 24 Aug. 2022

6. Marc Starnes, Large Truck Causation Study: An Initial Overview, FMCSA (March 2006).