A single calendar date for "fleet maintenance" is a schedule built for a fleet that doesn't exist — a real mixed fleet has as many service intervals as it has manufacturers. The honest version runs off each machine's own hour meter, not one date on the wall calendar.
Key takeaways
A yard with an excavator, a skid steer, two different generator models and a compressor, from four different manufacturers, doesn't have one maintenance interval — it has up to four, each set by that specific manufacturer's own service schedule and tied to hours of use, not the calendar. A single “maintenance day” scheduled once a quarter for the whole yard will be right for whichever machine happens to line up with that date and wrong for the rest.
The honest structure is per-machine: each unit's own hour-meter or odometer trigger, checked against its own manufacturer interval, logged against its own record — not one calendar event covering everything the yard owns.
Alberta’s OHS Code Part 2 requires an employer to “assess a work site and identify existing and potential hazards before work begins,” prepare a report of the results, and repeat the assessment “when a new work process is introduced,” when a process or operation changes, or before significant additions to a work site. A piece of equipment reaching the point where it needs maintenance it isn't getting is exactly the kind of change that reassessment is meant to catch — a maintenance schedule that flags overdue equipment is doing part of that statutory hazard-assessment work, not a separate task from it.
BC's framework runs on the same underlying logic even where the specific trigger differs — BC’s OHS Regulation Part 20 requires 24 hours' written notice for construction work meeting certain cost or structural thresholds, and separately assigns a qualified prime contractor role for coordinating hazards across a site. Equipment condition is squarely inside the hazard picture either framework is asking a business to manage, whichever province the work is in.
WSIB’s own page is explicit: “our premium rate-setting model has replaced our experience rating programs (NEER and CAD-7) and all programs concluded in fall 2020.” Anyone still describing WSIB's rates as running on NEER, CAD-7 or MAP is describing a retired system — the current model is different and worth understanding on its own terms, not through the old program names.
The current model, per WSIB’s current rate-setting guidance, sets rates “using your insurable earnings, claims costs and the number of allowed claims, over a six-year period,” assigning each business to a risk band where “the differences between each risk band rate is approximately five per cent.” A single serious claim traceable to equipment that wasn't maintained doesn't just cost the repair and the claim — it feeds a six-year window that can move the whole business into a costlier risk band, for every insurable dollar of payroll, for years after the incident itself is resolved.
Procore’s Equipment tool lets a company “perform equipment inspections” and “create and scan QR codes” tied to the specific machine's record, tracking it “from arrival to departure” — which is the structural answer to a mixed fleet: the trigger lives on the machine, not on a shared calendar. A paper system that tags each machine with its own manufacturer-set interval and hour-meter reading, checked at a fixed cadence rather than relying on someone remembering, gets most of the same result.
Skipped maintenance shows up as downtime before it ever shows up as a claim — a machine that fails mid-job because a scheduled service was skipped costs the repair, the rental of a replacement while it's down, and whatever the crew waiting on it was billing that day. That cost lands regardless of whether the failure ever becomes a WSIB claim, which is a reason to track PM compliance as its own metric rather than waiting to find out about a gap only after an incident forces the question.
A simple compliance rate — machines serviced on schedule divided by machines due for service in a given period — is enough to spot a fleet-wide slip before it turns into either a breakdown or a claim, and it's a number that can be checked monthly without much overhead once each machine's own interval is being tracked.
A worked example
A yard runs three machines: an excavator due for service every 250 hours, a compressor due every 500 hours, and a generator due every 400 hours. A single quarterly “maintenance day” scheduled every 13 weeks happens to line up reasonably well with the compressor (which logs roughly 480 hours a quarter at typical use) but badly with the excavator, which logs 300 hours in the same period — meaning it's overdue by 50 hours before the next scheduled maintenance day even arrives.
Tracked instead by each machine's own hour meter against its own manufacturer interval, the excavator gets flagged mid-quarter when it crosses 250 hours, not at the fixed calendar date — catching the overdue window the single shared schedule missed entirely.
No. WSIB’s own page confirms that those programs, along with MAP, concluded in fall 2020 and have been replaced by a different rate-setting model based on six years of claims history. Describing a current WSIB rate using the old program names is describing a system that no longer exists.
Alberta’s OHS Code requires it “at reasonably practicable intervals,” and specifically when a new work process is introduced, when a process or operation changes, or before significant site additions — equipment falling behind on its maintenance interval is a change worth triggering a reassessment over, not something to wait on until the next scheduled review.
By hours of use (or odometer, for road equipment), checked against each manufacturer's own interval — a calendar date works only for a fleet where every machine happens to log similar hours at a similar pace, which a genuinely mixed fleet rarely does.
Indirectly, over time — WSIB’s current model sets rates from six years of claims history, so fewer equipment-related incidents over that window supports a better risk-band placement. It isn't a direct, immediate discount for a maintenance record; it's the claims outcome that record is meant to prevent that eventually shows up in the rate.
A 30-minute call is enough to tell you whether AI pays for itself here.
