The Most secure Fleets Aren’t the Ones with the Most Expertise
You’ve made the telematics funding. Right here’s the right way to seize the opposite half of its worth and construct the most secure fleet on the highway.
Cameras seize occasions. Sensors detect harsh braking. Algorithms generate scores. All of it’s helpful. None of it, by itself, modifications how a driver drives.
I’ve spent years finding out why fleet security applications plateau. The reply will not be that fleets lack knowledge. Most are drowning in it. The reply is that knowledge, by itself, doesn’t change how a driver thinks. And till you modify how a driver thinks, you haven’t modified how a driver drives.
Right here is the strain price sitting with. Over the previous decade, fleets have spent closely on visibility, but the trade’s security numbers haven’t adopted. Between 2014 and 2022, as GPS telematics and video adoption climbed towards report ranges, the fatal-crash involvement price for big vans rose roughly 31 %, and the variety of individuals killed in crashes involving giant vans rose about 52 % (FMCSA and NHTSA). Shopping for the expertise has not, by itself, made fleets safer. The fleets that did get safer used the expertise for one thing extra.
The Proof Is Not Ambiguous
Each critical examine reaches a model of the identical conclusion: monitoring alone produces significant however restricted outcomes, and including structured behavioral teaching roughly doubles the end result.
One of many largest video telematics suppliers printed knowledge exhibiting that fleets utilizing cameras and telematics with out constant teaching noticed a 37 % discount in collisions, whereas fleets including constant teaching to the an identical expertise noticed a 73 % discount.
Placing apart the necessity to take these self-reported numbers with a grain of salt, this does present that even these expertise corporations’ personal numbers validate this truth. The expertise was the identical in each teams. The one variable was whether or not anybody used the information to coach and coach.
Peer-reviewed analysis confirms the extra modest baseline. A 2025 AAA Basis randomized managed trial discovered telematics suggestions diminished dashing 11–13 % and arduous braking 16–21 %, translating to roughly a 4–5 % discount in crash-claim frequency. An actual impact, and an order of magnitude beneath vendor advertising. Which makes the teaching hole—the gap between 37 and 73—an important quantity within the trade.
Peer-reviewed analysis confirms the extra modest baseline however tells the identical story. The NETS Energy in Numbers benchmark, aggregating security efficiency from dozens of enormous company fleets, confirms the sample:
- 3–4 proportion factors decrease collision involvement for fleets with closed-loop corrective-action follow-through
- About 3 proportion factors decrease for fleets utilizing in-person behind-the-wheel coaching versus distant or digital-only
- 20–30 % collision-rate discount for fleets combining telematics with energetic teaching applications
This makes the teaching hole—the gap between 37 and 73—an important quantity when contemplating the right way to get essentially the most worth out of your expertise funding.
Why “Extra Teaching” Is Not the Reply Both
At this level, the plain conclusion is: add teaching. That’s directionally proper. However the type of teaching issues enormously.
Most fleet teaching immediately works like this. A supervisor will get an alert. They sit down with the motive force and say, “You had 14 harsh-braking occasions this month. That should come down.” The motive force nods. They fight tougher for every week and enhance. However the occasions come again.
The issue is that the teaching will not be related to something the motive force was ever taught. The motive force hears a quantity, not a precept. They don’t be taught that harsh braking is often the downstream consequence of failing to scan the site visitors atmosphere far sufficient forward. That connection—from the occasion the expertise captured, to the choice that brought on it, to a precept the motive force discovered in coaching—is the through-line that separates applications that plateau from applications that maintain bettering.
The most recent model of this drawback wears a extra subtle face. Telematics and digital camera distributors are actually rolling out AI-generated teaching: automated suggestions movies, AI avatars that ship customized weekly debriefs, in-cab audio prompts, and machine-written speaking factors for managers. It’s a actual advance in effectivity. However it’s not an answer to the underlying drawback, as a result of it begins in precisely the identical place event-based methods at all times have: with a flagged occasion and a quantity on a dashboard.
AI teaching can inform a driver to cut back arduous braking or enhance a following-distance rating quicker and extra cheaply than an individual can. What it can not do, any higher than a rushed supervisor can, is join that correction to a driving precept the motive force was really educated in. The supply acquired cheaper. The mechanism of conduct change didn’t change.
A 2025 examine from the College of California, San Diego, funded by one of many video telematics distributors itself, concluded that manager-led, customized teaching stays the simplest technique for sturdy conduct change, and warned that automated teaching with out actual human interplay can cut back driver engagement and belief. Automating a disconnected dialog makes it quicker. It doesn’t make it work.
The fleets that break by the expertise ceiling share a standard attribute: they function from a named, structured behavioral framework. Not a algorithm. Not an inventory of violations. An outlined set of driving choices that drivers be taught, coaches reference, scorecards measure, and corrective motion reinforces—the identical language at each stage.
At Smith System, we’ve spent greater than 70 years constructing and validating that framework, the Smith5Keys. However the precept is greater than anybody supplier. Any fleet that desires to interrupt by the monitoring ceiling has to reply one query: does your program have a shared behavioral language that connects coaching, teaching, scoring, and corrective motion? If the reply is not any, the proof says you’re leaving half the end result on the desk.
What Occurs When the Loop Truly Closes
A nationwide meals and beverage retailer working 22 distribution facilities with greater than 2,200 drivers had telematics in place earlier than it related a behavioral framework to this system. After integrating behind-the-wheel coaching, behavioral scoring, and training referencing the identical driving choices, preventable accidents dropped 38 % over two years. Common indemnity value per incident fell from about $43,000 to $1,134.
The identical {hardware} was producing the identical indicators. What modified was that these indicators started flowing by a framework that translated them into particular, coachable choices.
Essentially the most sturdy proof comes from a state transportation division that adopted the identical framework by a peer-led, train-the-trainer mannequin. Annual incidents fell roughly 30 % within the first 12 months and held there because the workforce grew. Then this system was paused throughout the pandemic, and incidents climbed again to baseline. When coaching resumed, they dropped once more—and the fleet’s finest 12 months on report got here a number of years into this system, not at first.
There is no such thing as a novelty impact that lasts six years, disappears if you cease, and returns if you restart. That’s what a sustained behavioral system appears like: features compound when you preserve them, and reverse when you don’t.
5 Questions for Your Personal Program
No matter who constructed your program or what expertise you run, these questions will let you know whether or not you’re capturing the total end result or leaving outcomes on the desk:
- Can your teaching language hint again to one thing the motive force was educated in? Teaching that doesn’t reference coaching is a dialog about numbers. Teaching that reinforces coaching is a dialog about rules.
- Does your scorecard let you know which choice to educate, or simply which driver to fret about? If it flags a high-risk driver however can not let you know whether or not the issue is scanning, house administration, or situational consciousness, you have got a rating device, not a training device.
- When a driver completes corrective motion, do you confirm that the focused conduct really modified? Completion charges are an administrative metric, not a security metric. An open-loop system assumes coaching works with out ever checking.
- Are you able to determine the place threat is concentrating throughout your fleet by conduct—not simply by driver? A systemic following-distance drawback wants a unique response than a dashing drawback. Deal with the basis trigger at scale, not signs one driver at a time.
- Has your collision-rate enchancment continued previous the primary 12 months, or has it plateaued? Each monitoring program tends to have a powerful first 12 months. In case your enchancment curve has flattened, you have got doubtless reached the ceiling of what commentary alone can ship.
The Different Half
The fleet security trade has spent the final decade investing in visibility. However visibility was half of the issue. The opposite half—turning what the expertise sees into conduct change that compounds over time—requires a behavioral framework that drivers be taught, coaches reference, scorecards measure, and applications reinforce.
The most secure fleets aren’t those with essentially the most superior {hardware}. They’re those that constructed the bridge between what the expertise sees and what the motive force does. The proof on that time will not be ambiguous, and it’s not combined. The primary half of the funding has been made. The query for each fleet security chief is whether or not the second half is in place, or whether it is nonetheless ready to be constructed.
Concerning the Writer: Derek Dunaway is the Chief Govt Officer of Smith System, which has set the usual for driver security coaching for greater than 70 years. The corporate’s Smith5Keys® methodology is the behavioral basis of driver threat administration applications for among the largest and most secure fleets in North America.
Editor’s Word: This contributed article displays the creator’s opinion on the position of expertise, coaching, and training in fleet security and doesn’t essentially mirror the views of Automotive Fleet.
