For years, Fleet Managers have relied on telematics to identify speeding, harsh braking and other risky driving behaviours. The challenge has never been collecting data. The challenge has been knowing which data matters and how to act on it before an incident occurs.
According to Chris Martin, Senior Manager, Solutions Engineering – APAC at Geotab, artificial intelligence is changing that equation by shifting fleet safety from reactive reporting to predictive risk management.
Speaking with Fleet News Group, Martin explained that AI is enabling fleets to identify collision risks before they happen by analysing patterns across millions of data points that would be impossible for a human to detect.
“Fleet Managers have known for a long time that the data is key to decision making,” said Martin.
“We’re collecting vast quantities of it, and we’ve had really good analytics tools for many years, trying to summarise, aggregate, present insights in that data, but AI is shifting the needle in terms of how much data we can process, how quickly, and the types of patterns that we can find in data.”
Moving Beyond Traditional Safety Metrics
Traditional telematics programs often rely on predetermined thresholds such as speeding events, harsh braking or excessive acceleration. While these metrics remain valuable, Martin believes they only tell part of the story.
“If you look at a predictive safety perspective, we’ve been able to train a model based on real-world collision data to go back and look at historical driving patterns and say what are the characteristics or the driving behaviours that consistently precede a collision,” he said.
“These are the sort of patterns of data that a human can’t find.”
Rather than focusing on a single event, AI can assess combinations of variables such as vehicle type, location, operating conditions, climate, route characteristics and long-term driver behaviour.
“We might say a G-force value of point eight Gs might contribute to or might exhibit a good example of a harsh driving event, and we’re fairly confident that correlates to collisions,” Martin explained.
“But one event alone doesn’t necessarily give you a strong enough indication.”
The result is a more sophisticated view of risk that moves beyond simple event reporting.
Predicting the Probability of a Collision
One of the most significant developments is the ability to assign a predictive collision risk score to drivers and fleets.
Martin said Geotab’s AI models have been trained using real-world collision data and then tested against actual fleet operations.
“We’ve not only built the model but then tested it on real-world fleets, and we’re finding a very high accuracy of being able to identify a predictive collision risk based on those safety behaviours,” he said.
That allows Fleet Managers to receive insights that were previously unavailable.
“We might say, you’ve got an eight per cent risk of a collision in the next 100,000 kilometres. That’s a pretty useful insight.”
Rather than waiting for an incident to occur, fleets can proactively focus coaching, training and intervention efforts on drivers or operating groups with elevated risk profiles.
Benchmarking Against Similar Fleets
Predictive analytics becomes even more valuable when combined with benchmarking.
Martin noted that context is critical when evaluating safety performance. Comparing a metropolitan courier fleet with a long-haul transport operation would provide little value.
“There’s no point comparing a light ute with a heavy vehicle, or a long-haul transport company with a last-mile delivery operation,” he said.
By analysing similar fleets operating in comparable environments, AI can provide meaningful context around safety performance.
“We can give you a benchmark of whether your fleet is actually performing well from a safety perspective. Is eight per cent good or is eight per cent bad?”
For Fleet Managers, that creates a clearer understanding of where improvement opportunities exist and whether current safety initiatives are delivering results.
Reducing the Workload for Fleet Managers
The benefits extend beyond safety outcomes.
Many Fleet Managers are responsible for hundreds or thousands of vehicles while also managing procurement, maintenance, compliance and operational demands. AI can reduce the burden of analysing large volumes of data and configuring complex reports.
“This all happens with AI driving it without having to run the reports, tweak the reports, work out your own risk thresholds,” Martin said. “It significantly reduces the workload for the fleet manager to give them those insights.”
The introduction of tools such as Geotab Ace is further simplifying access to information by allowing users to ask questions in natural language.
“You can just say, ‘Hey Geotab, who are my top 10 most risky drivers currently?’ and you can get that response straight away,” said Martin.
Fleet Managers no longer need to be experts in report building or data analysis to gain valuable operational insights.
The Future of Fleet Safety
As AI models continue to evolve and learn from larger datasets, Martin expects predictive analytics to become increasingly important across fleet operations.
While predictive safety is one of the most immediate applications, similar approaches are already being applied to maintenance, vehicle uptime and electric vehicle performance.
“There’s always more innovation to do,” Martin said.
“As AI models improve and the models learn more about real-world behaviours, we’re going to see constant improvements and new features brought into the Geotab platform.”
For fleet operators, the shift represents a fundamental change in how telematics data is used. Instead of analysing what happened yesterday, AI is helping fleet managers understand what is likely to happen tomorrow — and giving them an opportunity to act before it does.
- Pajero Is Back: Mitsubishi Revives an Australian 4WD Icon
Mitsubishi has confirmed the Australian range and specifications for the all-new Pajero, with the returning nameplate bringing a 2.4-litre turbo-diesel engine, seven drive modes, a 3.5-tonne towing capacity and a choice of five- and seven-seat configurations. The Mitsubishi Pajero will return to Australian showrooms in December 2026, reviving one of the brand’s best-known nameplates after - Jaecoo Strengthens the J7 Line-UpJaecoo Australia is rolling out its MY27 J7 range, with the midsize SUV continuing to offer both petrol and plug-in hybrid powertrains as the brand broadens its appeal to Australian buyers. The updated line-up includes the petrol-powered Ridge AWD alongside three Super Hybrid System – PHEV variants: the Track SHS-P, Summit SHS-P 2WD and Summit
- Nissan Patrol Enters a New EraNissan has confirmed key Australian specifications for the new-generation Y63 Patrol, with the large SUV gaining a new twin-turbo V6, increased towing capacity, a revised chassis and a broader six-grade line-up. The new Patrol will arrive in Australia in early 2027, replacing the long-serving Y62 and introducing significant changes to the powertrain, suspension, cabin technology
- Hyundai Gives Palisade More Off-Road BiteHyundai will add a new Palisade XRT Pro variant to its Australian range from September, giving the large SUV more ground clearance, all-terrain tyres and a locally developed suspension tune. The XRT Pro is Hyundai Australia’s first model to use the XRT Pro designation. It sits alongside the existing Palisade Elite and Calligraphy variants, but differs through
- Mercedes-Benz Sharpens the C-ClassMercedes-Benz has unveiled a heavily upgraded C-Class Saloon and Estate, bringing new electrified powertrains, artificial intelligence, improved driver assistance systems and greater digital functionality to one of the brand’s most established fleet models. The update is described by Mercedes-Benz as the most comprehensive technical upgrade in the history of the C-Class, with changes spanning design,
