Utilization is the share of the time a machine could have worked that it actually did work. The math fits on a napkin. The arguments start over what counts as “could have worked” and what counts as “worked,” and those two choices decide whether the number helps you sell a loader or just decorates a report.
This guide gives you the formulas, a worked example with round numbers, a way to set a benchmark from your own costs, and the mistakes that quietly skew the result.
The basic formula
Time utilization = engine hours ÷ available hours × 100
Engine hours are the hours the engine ran in the period. Available hours are the hours the machine was expected to be ready for work: working days in the period times the scheduled shift length.
Do not use calendar hours. A machine on one 10-hour day shift, five days a week, is scheduled for about 220 hours in a month of 720, so a calendar denominator caps it near 30% even when it never stops during the shift. Every machine looks idle, and the ones that really are idle disappear in the crowd. Use the schedule your crews actually work, and change it per job when a job runs two shifts.
Three rates worth tracking
One number cannot answer every question, so keep three side by side.
- Time utilization. Engine hours ÷ available hours. It answers “is this machine running when we need it?”
- Working utilization. (Engine hours minus idle hours) ÷ available hours. It answers “is it doing work when it runs?” Idle hours come from the engine controller, which logs time at low idle with no load.
- Dollar utilization. The value of hours charged to jobs ÷ the machine’s annual owning cost. It answers “does this machine earn its keep?” Rental companies use a version based on rental revenue and original cost; a contractor can use internal equipment rates instead.
A fourth figure, idle share (idle hours ÷ engine hours), is not a utilization rate. It explains the gap between the first two, and it is often the cheapest gap to close.
Example: one excavator, one month
Take a 30-ton excavator on a road job. These are round numbers for illustration, not a benchmark:
- Working days in the month: 22
- Scheduled shift: 10 hours
- Available hours: 22 × 10 = 220
- Engine hours from the engine controller: 154
- Idle hours from the engine controller: 44
Time utilization is 154 ÷ 220 = 70%. Working utilization is (154 minus 44) ÷ 220, which is 110 ÷ 220 = 50%. Idle share is 44 ÷ 154, about 29%.
On a whiteboard, 70% reads as a busy machine. The working figure says half the shift produced work, and almost a third of the engine time burned fuel and service hours at idle. Both are true. The first number tells you not to pull this excavator off the job. The second tells you to ask the superintendent about truck spacing, long warm-ups or an operator who leaves it running through lunch.
Those 44 idle hours also count toward the next oil change. On a schedule keyed to engine hours, idle time pulls every service closer without moving any dirt.
Rolling up a fleet honestly
Fleet utilization is the sum of engine hours divided by the sum of available hours, calculated within one machine class. Averaging the percentages of a dozer and a light tower tells you nothing about either.
Here are five wheel loaders, each with 200 available hours last month:
| Loader | Engine hours | Time utilization |
|---|---|---|
| L-1 | 180 | 90% |
| L-2 | 170 | 85% |
| L-3 | 150 | 75% |
| L-4 | 30 | 15% |
| L-5 | no data | unknown |
If L-5 really sat all month, the class rate is 530 ÷ 1,000 = 53%. That average describes no actual loader. Three are busy, one is barely used and one is a blank. The decision lives in the spread: is L-4 surplus, and where is L-5?
A machine that reports nothing is not at 0%. It is unknown, and it belongs on its own list until someone confirms it is parked, broken or gone. Leaving it out raises the average; counting it as zero lowers it. Neither is ground truth.
A benchmark from your own costs
Published targets float around, but the useful benchmark comes from your own books: the utilization at which owning a machine beats renting one.
Break-even hours per year = annual owning cost ÷ rental cost per hour
Break-even utilization = break-even hours ÷ available hours per year
Example: a wheel loader costs $50,000 a year to own once you add depreciation, interest, insurance and storage. Say a comparable rental runs $8,000 per four-week cycle, which covers 160 hours, so $50 an hour. Break-even is 50,000 ÷ 50 = 1,000 hours a year. With 250 working days of 10 hours, available hours are 2,500, so break-even utilization is 40%.
Leave fuel and the operator out of both sides, since you pay them either way. Above 40%, owning is the cheaper way to have that loader. Below it, for a sustained stretch, renting would have cost less. That benchmark is tied to money, and it is different for every class.
Plug your own figures into the equipment tracking ROI calculator, and see our rent or own decision rule for the full version with haul and rental terms included.
Track availability beside it
Utilization asks whether a machine worked. Availability asks whether it could have.
Availability = (scheduled hours minus downtime hours) ÷ scheduled hours × 100
Say another excavator was scheduled for 220 hours and spent 22 of them waiting on a hydraulic hose and a mechanic. Availability is (220 minus 22) ÷ 220 = 90%. If you leave downtime inside the utilization denominator, a machine that sat broken for a week looks like a machine nobody wanted. Keep the two numbers apart and each one points at a different fix: the shop for availability, dispatch for utilization.
Decision rules you can use
Write these down with your own thresholds, then apply them every month instead of debating each machine from memory.
- Under benchmark three months running, with no seasonal reason: offer it to another job, rent it out, or put it on the sell list.
- Above a high mark you set (say 85%) for a quarter, in a class where you also rent: price one more owned unit against the rental spend.
- Idle share above a third: check idle shutdown settings, warm-up habits and truck spacing before blaming the machine.
- Hours logged with no job assigned: fix the assignment. Unassigned hours cannot be costed or billed.
- No report for seven days: treat the machine as unknown and find it this week.
The rules matter more than the exact numbers in them. A written rule turns a monthly argument into a monthly checklist.
Common mistakes that skew the rate
- Calendar hours in the denominator. Every machine looks idle, so nothing stands out.
- Counting idle as work. Report time and working utilization together, always.
- Dropping machines that stopped reporting. That inflates the average and hides the real problem.
- Hand-copied hour meters. Monthly readings drift, and a replaced gauge cluster can reset the count. Use engine controller hours where you can.
- Treating shop time inconsistently. Decide once whether a machine down for repair counts as available. The cleaner habit is to exclude it and report availability as its own number, as above.
- Key-on time instead of engine hours. A tracker that only sees ignition voltage counts a key left on at lunch as work.
- Averaging across classes. Roll up within a class, then compare each class to its own benchmark.
- Ignoring attachments. A breaker or a compaction wheel can sit for months while its carrier looks busy. Tag the attachment and measure it too.
Clean hours every Monday
The formula only works on hours you trust. The best source is the engine controller. An A1 gateway reads J1939 engine hours and idle time directly, and newer machines report through their manufacturer’s feed under ISO 15143-3, which our guide to AEMP 2.0 feeds explains. An hour meter photo at the pre-shift inspection is a fallback, not a plan.
A short Monday routine:
- Pull last week’s engine and idle hours by machine and by job.
- Sort each class from lowest to highest time utilization.
- List machines under benchmark and anything not reporting.
- Check open rentals in the same classes as your least-used owned machines.
- Send the superintendent one line per decision: move, release or keep.
It takes about twenty minutes, and it replaces a lot of phone calls.
Where Gytix fits
Utilization and cost analytics runs this math every day for every machine, owned or rented. It shows time and working utilization, idle share, class rollups and a list of machines that stopped reporting, with the source of every hour figure. Utilization reports are part of the Field plan.
The same hours feed job costing, right-sizing and the AI maintenance agent inside fleet intelligence, so one clean hour count serves the shop, the office and the superintendent.
