Fleet utilization in construction is usually reported as one number, but it is really four different measurements, and they need different data to produce. Time utilization (how many days an asset was assigned to a job) can be calculated from scheduling and trip records alone. Engine utilization (how many hours the machine actually ran, and how much of that was idling) cannot: it requires a reading from the machine itself, through an ECU port, an OEM telematics feed or a fitted hour-meter sensor. Most disputes about utilization numbers come from comparing one of these to the other.

This article sets out the four fleet utilization metrics used in construction, the three ways to measure them, and how to tell which one a vendor is actually offering you.

What fleet utilization actually means

The word is used for at least four distinct ratios. Agreeing on which one you mean is the first step, because a fleet can look 85% utilized on one definition and 31% on another in the same month.

The four metrics

Time utilization — was the asset committed to a job? Formula: days assigned ÷ days available.

Engine utilization — did the machine actually run? Formula: engine hours ÷ hours available.

Productive utilization — was the machine working, or running and standing still? Formula: (engine hours − idle hours) ÷ engine hours.

Financial utilization — did the asset earn against its cost? Formula: revenue or internal hire value ÷ ownership cost.

A common failure: a fleet reports 85% because the excavator was allocated to a site for 26 of 31 days, while the machine logged 48 engine hours in a month with roughly 250 working hours available. Both numbers are correct. They answer different questions, and only the second one tells you whether to move the machine or sell it.

Three ways to measure fleet utilization

Machine telematics — ECU port, OEM feed or aftermarket box

• Reads: engine on/off, hour meter, fuel burn, fault codes, sometimes load
• Gives you: engine hours, idle time, fuel per hour, usage-based service triggers
• Cannot tell you: who was operating, why the machine sat, what the job was worth
• Hardware: required, fitted per machine or bought as an OEM data subscription

Asset GPS tracker — battery or wired unit

Reads: position over time

• Gives you: location history, movement versus stationary, geofence entry and exit, theft alerting

• Cannot tell you: whether a stationary machine is idling or switched off

• Hardware: required, one unit per asset

Phone-based trip and assignment records

Reads: operator check-in, trip start and end, arrival and departure at a geofenced site, task completion

• Gives you: time utilization, site attendance, travel versus on-site hours, assignment history, proof of work

• Cannot tell you: engine hours, idle hours or fuel burn — none of these can be derived from a phone

• Hardware: none, runs on the operator's existing phone

This is the line that matters. A phone reports the position and activity of the phone, and by extension the person carrying it. It is a reliable record of who was assigned where and for how long. It is not a reading from the machine, and no amount of modelling turns an arrival time into an engine hour. A vendor claiming engine hours or predictive maintenance from a phone alone is describing an estimate, and should be asked to say so. Where a machine already reports to a telematics or OEM system, that feed is a second, independent data source — and the two can be read side by side, which is how most mixed fleets end up with a complete picture.

Worked example: one excavator, three utilization figures

One 20-tonne excavator, one calendar month, 22 working days.

Scheduling records — assigned to a site on 19 of 22 days. Result: 86% time utilization.

Asset GPS — stationary within the site geofence on 19 days, moved between sites twice. Result: confirms presence, no engine insight.

ECU hour meter — 31 engine hours logged, 9 of them below working load. Result: 18% engine utilization, 71% productive.

All three are accurate. If the question is "do we need to hire a second excavator next month", the 86% figure says yes and the 31-hour figure says almost certainly not. This is the single most expensive measurement error in construction fleets.

Which utilization metric fits which decision

You do not need engine-level data for every decision. Matching the method to the decision keeps the cost proportionate.

Phone-based assignment and site records are enough for:

Is this machine on the right site? — assignment and arrival records

Who was operating it, and did the task get done? — operator check-in and task completion

Should we bill this site for the machine's time? — assignment plus arrival and departure times

Engine data is required for:

Is this machine earning its keep? — engine hours

Are operators leaving machines idling? — engine hours plus idle split

When is this machine due for service? — engine hours, or a documented proxy, unless you accept calendar-based servicing

Should we buy, hire or dispose? — engine hours plus cost data

A practical sequence for a mixed fleet: start with phone-based assignment and site records across every asset, because the cost is near zero and it closes the "where is it and who had it" gap immediately. Then connect whatever telematics you already have, and fit new hardware only to the subset of machines where engine hours actually change a decision — usually the highest-value or highest-idle equipment, not the whole fleet.

If you already have telematics

Many fleets do, and never use it. Machines under warranty often report to an OEM portal nobody logs into, and aftermarket boxes fitted years ago frequently still transmit. Before buying anything, find out which of your machines already produce a feed and whether you have access rights to it.

Connecting an existing feed to the system that holds your trip and assignment records closes the engine-hours gap without fitting a single new device. It also solves the more common complaint, which is not that the data is missing but that it sits in a separate portal from the operational record, so nobody reconciles the two.

What to ask a fleet utilization vendor

These questions separate measured data from modelled estimates. They are worth asking of every supplier, including us.

Where does each number come from? Ask for the data source of every metric on the dashboard: machine, tracker, phone or manual entry.

Is "engine hours" measured or estimated? If estimated, ask what the estimate is built from and what its error range is.

What happens when the operator's phone is off? Ask how the record is completed, and whether the gap is visible in the report.

Which machines in my fleet can provide an ECU feed? OEM coverage varies sharply by brand, model and year. Ask for a per-model answer, not a general yes.

Is the OEM data feed included, or a separate subscription? For many brands the data is licensed separately from the machine.

Can I export the raw events? If you cannot export arrival, departure and engine events, you cannot audit the ratio or move it into your own reporting.

What is the definition behind each percentage? Ask the vendor to write the formula. If they cannot, the number is not comparable to anything else you measure.

Frequently asked questions

Can you track equipment utilization without installing hardware?

Partly. Without hardware you can measure time utilization, site attendance, assignment history and operator activity from a phone. You cannot measure engine hours, idle time or fuel burn — those require a reading from the machine. If your machines already report to an OEM or aftermarket telematics system, that feed can be connected and read next to the phone records without fitting anything new.

What is a good utilization rate for construction equipment?

There is no single benchmark, because the figure depends on which of the four definitions you use and on the asset class. A support machine that must be available on demand will show low engine utilization by design. Compare a machine to its own trend and to similar assets in your fleet, not to a published industry figure.

What is the difference between engine hours and operating hours?

Engine hours count every hour the engine runs, idling included. Operating hours, sometimes called productive hours, exclude idle time. The gap between them is usually where the recoverable cost sits.

Do I need a tracker on every machine?

No. Fit hardware where engine data changes a decision — high-value assets, machines suspected of heavy idling, and anything on usage-based servicing. For the rest, assignment and site records are usually enough.

How is idle time calculated?

From the machine: the engine is running while the machine registers no working load or movement for a set period. It cannot be inferred from a phone sitting still, because a stationary phone is equally consistent with a switched-off machine.

Can utilization data support usage-based maintenance scheduling?

Only if the hours come from the machine, which an existing telematics feed satisfies and an estimate does not. Service intervals built on estimated hours drift from the real figure and either over-service or miss the interval. If you do not have engine data, calendar-based servicing with a documented assumption is the honest alternative.

Related: Trailer tracking without a device on every trailer

Cooriroo is an AI operations platform for fleet, workforce and field operations. It records assignments, trips, site arrivals and proof of work from the operator's phone, and connects to the telematics or OEM systems you already run so that machine-level data appears alongside them. Cooriroo does not supply tracking hardware. If you want to work out which parts of your fleet need machine data and which do not, book a 15-minute call.