Rising fuel prices are putting pressure on fleet operating budgets, but they may also be exposing a problem many organisations don’t realise they have: fuel theft.
In a discussion with fleet technology specialists Webfleet and Roev at Sustainability Business Live 2026, advances in telematics and artificial intelligence are making it easier to identify fuel anomalies that previously went undetected, even in large fleets with sophisticated reporting systems.
The issue came to light through a recent customer case where Roev’s analytics platform identified suspicious fuel activity that had not been detected through traditional reviews.
“There’s a few ways we’ve been able to help them, one of which is in anomaly detection,” said Robert Dietz, Co-founder and CEO at Roev.
“There’s a number of things you can basically look for, you know, over tank fillings or just things that are out of the norm. When you double click on it and use the telemetry to back it up, you can actually see that the vehicle is running at the petrol station for 90 minutes and then all of a sudden makes no sense why that fueling is occurring.”
The customer operated a fleet of approximately 300 vehicles nationally. According to Adam Dennis, Enterprise Account Manager at Webfleet, the suspicious activity was only uncovered after Roev’s software analysed the data and highlighted unusual patterns.
“In this particular example they had been reviewed by a major consulting firm, and no anomalies were found,” Dennis explained.
“But Robert found about 30 anomalies, a couple of which were like $1,000 in one refuelling event. We were questioned on it, and obviously we had to prove it, and we found that the vehicle was parked up for 90 minutes idling at a petrol station.”
Dennis said the combination of Roev’s analytics and Webfleet’s telematics data provided the evidence required to investigate the issue.
“The Roev data highlighted the anomaly and then we were able to prove it,” he said.
Beyond obvious examples of fuel misuse, the technology is also capable of identifying more subtle behaviours that may warrant investigation.
Dietz explained that unusual weekend fuel purchases, unexpected refuelling frequency or fuel transactions occurring outside normal operating areas can all be indicators that something deserves closer attention.
One Victorian council, for example, discovered that a fuel card assigned to one vehicle was being used elsewhere after visualising fuel transaction locations on a map.
“There was a fueling that was very clearly outside their service area,” Dietz said.
“What they found is the card was leaving the vehicle it was assigned to and actually going with a different vehicle and being used to fuel up in that area.”
The ability to visualise data is proving just as valuable as the analytics itself.
Rather than reviewing spreadsheets containing thousands of transactions from multiple fuel card providers, fleet managers can quickly identify unusual patterns and focus their attention where it is needed.
Dietz stressed that the software does not determine whether fraud has occurred.
“We detect the anomalies, we don’t judge them,” he said.
“We just call them out. In some cases there’ll be a difference in the use case. Maybe the vehicle had an extra tank on the back or was pulling a lawn mower. We don’t have an opinion on it, we just elevate them.”
Importantly, anomalies are not always linked to employee misconduct.
Unexpected fuel consumption can also indicate mechanical issues, fuel leaks or deteriorating vehicle performance.
“If you do start seeing strong deltas in fuel usage and start calling them out, it could actually lead to something physical with the vehicle,” Dietz said.
“There could be a leak or there could be a major incident going on with the vehicle, so its efficiencies dropped dramatically.”
The discussion comes at a time when cost-of-living pressures are increasing across Australia, creating conditions that may contribute to fuel misuse.
Scott Elkington, Marketing Manager at Webfleet, pointed to the classic fraud triangle of pressure, opportunity and rationalisation.
“People’s needs change. We hear it in the media, like cost of living pressure, and fuel’s obviously a very large part of that,” Elkington said.
“They can rationalise it. They think the business doesn’t really notice that much fuel out of how much they use. It’s only a little bit of money.”
While a single fuel transaction may seem insignificant, the cumulative impact across a large fleet can become substantial.
The data suggests the issue may already be growing.
“We are seeing an increase in anomaly detection of about 28 to 31 per cent,” said Dietz.
While not every anomaly represents fraud, the trend highlights the growing value of combining telematics data with advanced analytics.
For Fleet Managers, the lesson is simple. Fuel cards, telematics and transaction records already generate enormous amounts of data. The challenge is no longer collecting information — it’s identifying the small number of events that deserve attention.
As fuel costs continue to rise, AI-powered anomaly detection may become one of the most effective tools fleets have to protect their operating budgets.
