The intersection of AI, telematics, and data analytics is transforming road safety in ways never seen before. At Geotab Connect 2025, Mike Branch, Vice President of Data & Analytics at Geotab, shared his insights on how predictive models are reshaping fleet safety, why data-driven decision-making is essential, and how fleet managers can leverage technology to reduce collisions and improve driver behaviour.
From Traditional Safety Metrics to AI-Powered Predictions
Historically, fleet safety programs relied on basic telematics data such as harsh braking, speeding, and acceleration events. While these factors provided a general indication of driver risk, they did not accurately predict the likelihood of a collision. Mike Branch highlighted the limitations of traditional driver scorecards, which typically assigned arbitrary weightings to various driving behaviours.
“A fleet manager doesn’t know whether to weight something 10% for harsh braking or 30% for speeding,” Branch explained. “Our AI-driven model eliminates that guesswork by analysing actual collision data and scientifically determining the risk factors that matter most.”
Fleet managers have long struggled with determining how to prioritise different risk factors. Some might believe that harsh braking is a more serious issue than speeding, while others may focus on acceleration. However, these weightings were often subjective and did not necessarily reflect actual risk. AI has removed this guesswork by training models on real collision data, allowing fleets to understand the true correlation between driving behaviours and accident probability.
The Power of Predictive Collision Risk Models
One of the key advancements Geotab has introduced is a predictive collision risk model that goes beyond simple event tracking. By analysing extensive telematics data combined with real-world crash data, this model determines how specific driving behaviours impact the likelihood of an accident over a given distance, such as 100,000 miles. This approach not only improves accuracy but also provides fleet managers with clear, actionable insights to reduce risk proactively.
“I was at a conference speaking a couple of months ago about the new predictive collision risk models, and at the very end of it, I had a fellow come up to me who manages a fleet of multiple thousands. He said, ‘You mean, I don’t have to worry about weights anymore, right?’” Branch recalled. “That’s the tricky thing—fleet managers no longer have to guess how to prioritise different driving behaviours because the model figures it out for them.”
This shift allows fleets to move beyond subjective safety assessments and instead rely on concrete, data-driven risk analysis. The AI model continuously improves as more data is collected, allowing it to refine its predictions and provide even greater accuracy over time.
Integrating AI into Fleet Safety Systems
Geotab’s AI-powered safety system integrates additional contextual data, including following distance, weather conditions, and road type, to refine its predictions. Unlike older methods that assessed drivers in isolation, this new system benchmarks them against similar drivers operating under comparable conditions. This ensures that safety assessments are fair and relevant, preventing misclassifications that could arise from differences in geography or driving environments.
Branch explained how the AI model expands over time, incorporating new data points as they become available. “We have the model sitting there already. You don’t have to worry about all these weights, and you’re feeding more and more insight into this,” he said. “So now, from a product perspective, not only do you have this risk analytics portfolio in the Geotab platform, but as you look at further integration with OEMs, those safety features now all of a sudden become available in the in-vehicle infotainment (IVI) systems as well.”
Another major advantage of AI-driven safety analytics is its scalability. As more data is fed into the system, it continuously improves, making predictions even more precise over time. With the addition of OEM-integrated telematics and camera-based monitoring, fleet managers will have access to even richer data sources, allowing them to implement proactive safety measures rather than relying solely on post-incident analysis.
Why Fleets Need to Act Now
According to Branch, safety has now become the top priority for many fleets, surpassing even fuel efficiency and maintenance in importance. The financial impact of accidents—including vehicle repairs, downtime, legal liabilities, and reputational damage—has made safety an essential focus for fleet operators.
“If I survey our customers now, safety is the number one issue.”
Mike Branch
“A big part is, of course, that human lives matter, but also the cost to the company of accidents is so high when you look at the ancillary costs—not just vehicle repairs, but also having vehicles on standby, dealing with lawsuits, and other financial impacts.”
He also emphasised the importance of driver engagement. Instead of merely monitoring driver behaviour, companies need to create a culture where drivers actively participate in safety initiatives. By using AI to provide personalised insights and structured coaching, fleets can shift from punitive measures to a more collaborative approach that encourages positive behavioural change.
“It’s no longer just ‘my boss told me to do so,’” Branch said. “There’s something on the other end—a tangible benefit that drivers can work toward. It’s about leading with the carrot, not the stick.”
The Future of AI in Fleet Safety
Looking ahead, Geotab sees AI-driven safety analytics becoming standard across all fleet types, including passenger, light commercial, and heavy-duty trucking. As the technology evolves, the focus will expand beyond just accident prevention to include broader well-being initiatives for drivers.
“The joint venture with Vitality also operates on that same predictive collision risk model,” Branch noted. “As you do better from a predictive collision risk perspective, you get rewarded and incentivised based on the output of that model.”
Ultimately, fleet managers who embrace AI-driven safety analytics will be better equipped to prevent accidents, reduce costs, and create safer roads for everyone. As Branch concluded, “The data is already available—now it’s time for fleets to put it to work.”
