Introduction
Ask any owner of a service business with vehicles where the money goes, and the same two answers come back: fuel and repairs. A fleet of eight or ten vans can easily burn through $12,000 a month at the pumps, and one transmission failure at the wrong moment costs the repair bill plus a day of cancelled jobs and an unhappy customer. For years, small fleets simply absorbed these costs because the tools used by big logistics companies were priced for hundreds of trucks, not a dozen.
That has changed. In 2026, AI-powered fleet management has come down to a price point, often $30 to $50 per vehicle per month, that makes sense for a plumbing company in Trois-Rivières or a pest control outfit in Brandon. The technology reads engine data, driving patterns, and job schedules, then turns that stream into savings that show up within weeks, not quarters.
Where Fleet Money Actually Leaks
Most owners can tell you their monthly fuel bill but not their cost per kilometre per van, and that gap is where money hides. Two identical vehicles on similar routes can differ by 20 percent in fuel burn because of idling habits, route choices, tire pressure, and driving style. Multiply that gap across a fleet and a full year, and a business running ten vehicles can be overspending by $15,000 or more without ever seeing a number that looks wrong.
Maintenance is the second leak. The traditional approach is reactive, fix it when it breaks, or calendar-based, change the oil every 5,000 kilometres whether it needs it or not. Reactive maintenance means roadside callouts at $400 and up before a tow is even involved. Calendar maintenance means paying for service intervals the vehicle does not need while missing the ones it does, because a van doing short urban trips wears very differently than one running highway kilometres.
Predictive Maintenance Beats the Breakdown
Modern telematics hardware plugs into the vehicle's diagnostic port and reports continuously: battery voltage trends, coolant temperature behaviour, brake wear indicators, fault codes that have not yet triggered a dashboard light. AI models compare these signals against failure patterns from thousands of similar vehicles and flag problems early, often two to four weeks before a breakdown.
The economics are straightforward. Catching a failing alternator on a Tuesday afternoon costs $350 and an hour of downtime. Discovering it at 7 a.m. on a job site across town costs the part, a tow, a lost day of bookings, and a customer who may not call again. A nine-van plumbing fleet near Trois-Rivières reported cutting unplanned downtime by roughly 70 percent in its first year on predictive maintenance, which mattered more to them than the direct repair savings.
Routing and Idling: The Fuel Math
AI dispatch routing is where the fuel savings compound. Instead of technicians driving in the order jobs were booked, the system sequences calls by geography, traffic, appointment windows, and even which van has the right parts on board. Fleets commonly see 10 to 15 percent fewer kilometres driven for the same number of completed jobs, which at current prices is $1,000 or more per van per year.
Idling is the quieter target. A service van idling through winter warm-ups and long service calls can burn a litre or more per hour of stationary time, and across a fleet that adds up to thousands of dollars a season. Fleet platforms surface the worst offenders by vehicle, not by accusation, so managers can coach with facts. Most businesses find the biggest drop comes simply from making idling visible; nobody wants to top the chart two months in a row.
Driving Data Without the Big Brother Problem
The hesitation many owners feel is cultural, not technical. Technicians reasonably worry that tracking means surveillance. The businesses that roll this out successfully frame the data around the vehicle, not the driver: harsh braking events are flagged as coaching moments, not discipline, and location data is used for dispatch and customer ETAs rather than checking up on lunch breaks. Several Canadian fleets also give drivers access to their own scores, which turns the whole thing into something closer to a fitness tracker than an ankle monitor.
There is also an insurance angle worth checking. A growing number of Canadian commercial insurers offer meaningful premiums discounts, sometimes 10 percent or more, for fleets with verified telematics programs, which can cover much of the subscription cost on its own.
Conclusion
A service fleet will never be free to run, but it no longer has to be a black box with a fuel card attached. Small Canadian fleets using AI-driven management are finding four-figure annual savings per vehicle, fewer Friday-afternoon breakdown calls, and customers who get accurate arrival windows instead of "sometime between noon and five." The vans are already generating the data every kilometre they drive. The only real decision is whether you look at it.