Hour meter logging SOP for crews
A copy-paste SOP for daily hour-meter logging. Who logs, when, what to do when the meter is broken, and how to flag a number that doesn't match yesterday's.
Every fleet that logs hours well has a written SOP for it. Every fleet that doesn't, doesn't. This is a copy-paste standard operating procedure for end-of-shift hour-meter logging. It fits on a page, it answers the questions that come up in real crews, and it gives the shop foreman a tool to point at when an operator asks “do I have to do this every day.” The answer is yes, and here is exactly how.
Scope
Applies to every operator-rated asset in the fleet with a functioning hour meter: excavators, loaders, dozers, skid steers, articulated trucks, telehandlers, and any on-road truck that the shop tracks by hours rather than miles. Does not apply to trailers, gensets in standby, or hand tools.
Who logs
The operator of record for the shift. If a single operator ran the unit for the whole shift, that operator logs once at end-of-shift. If two operators ran the unit (a hand-off mid-day), each operator logs their own segment with a hand-off log at the swap point.
The foreman does not log on behalf of operators. The foreman reviews. Letting the foreman log creates the exact estimation problem the SOP exists to prevent — the foreman is not the person who watched the meter all day.
When to log
End of shift. Before clocking out. No exceptions.
The operator is standing at the unit. The engine is off. The meter face is visible. The phone is in the pocket. This is the only moment of the day when the cost of logging is zero seconds of mental overhead beyond the 20 seconds of actual taps. Any log entered later — at home, in the morning, on Friday — comes from memory and degrades to an estimate.
How to log
Two paths, in priority order:
- Three taps in the app. Open DirtFleet, pick the unit you ran (the app remembers your last asset; usually one tap), type the meter reading, submit. Total: 15–25 seconds with normal signal.
- Photo of the meter if signal is bad. Same screen has a camera button. Frame the meter face and the unit number-plate together if you can. Submit. The app queues it; OCR reads the digits when the photo lands on the server. Hour logs work offline; the queue ships when the phone next sees a cell tower (see Offline-first for the yard).
Fallbacks
Things go wrong in the field. The SOP covers the common cases.
Meter is broken or unreadable
Open the log screen. Toggle the asset to BROKEN METER. Estimate the runtime in hours for the shift you just worked. Submit. The log gets tagged ESTIMATED automatically and pulled out of the cost-per-hour math until the meter is fixed. Tell the shop foreman so the meter goes onto the next work-order list.
If the meter face is just dirty, wipe it and try again before falling back to estimation.
Missed shift
If a shift went unlogged — operator forgot, phone dead, unit ran without a logged operator — the next operator to use the unit logs both readings at their end-of-shift:
- Today's actual meter reading at end-of-shift.
- A second “catch-up” log for yesterday with the best available estimate (often: today's reading minus today's known runtime). Mark it BACKDATED in the note field.
The manager sees BACKDATED in the audit trail and knows the number is reconstructed, not direct.
Hand-off mid-shift
Outgoing operator reads the meter and submits a log. Incoming operator opens the same unit; the app pre-fills the hand-off reading as their starting point. At end of incoming operator's segment, they read the meter again and submit. Two logs, one unit, full attribution.
Anomaly handling
Two specific cases the SOP asks operators to flag instead of ignore.
- Today's reading is lower than yesterday's. The meter rolled over, was reset by a mechanic, or was replaced. Submit the reading anyway and add “meter appears reset/replaced” in the note. The flag goes to the mechanic queue automatically. Do not adjust your reading to make it look right.
- Today's reading is way higher than expected. More than 24 hours of runtime in a single shift, for example. Could be a fat-finger (typed 9999 instead of 1999) or unauthorized use. Submit the real reading; the anomaly detector flags it for review (see Spotting stuck meters and tool theft). The operator's only job is to type what they see.
Manager weekly review
Once a week, the manager opens the Hours dashboard and looks for three things:
- Assets with no log in the last seven days. Either out of service or being missed. Either way, action.
- Logs tagged ESTIMATED or BACKDATED. Should be a small fraction. If a particular operator or unit shows up here consistently, the SOP is not landing for that crew — usually a phone or login problem.
- Anomaly flags from the previous week. Resolve, dismiss, or escalate to the mechanic.
Fifteen minutes per week. Catches almost everything that goes sideways before it costs money.
POST THIS ON THE TOOLBOX LID
Hour meter logging — daily SOP
- End of every shift, before you clock out, read your meter.
- Open DirtFleet. Three taps. Submit. (Photo if signal is bad.)
- Meter broken? Toggle BROKEN METER and estimate. Tell the foreman.
- Handed off mid-shift? Each operator logs their own segment.
- Reading lower than yesterday? Submit anyway, add note. Flag for mechanic.
- Missed a day? Log today and a BACKDATED log for yesterday with best estimate.
One log, one operator, one shift. Before anyone leaves the yard.
Why this works
The SOP is short on purpose. Long SOPs do not get read. The rules are designed around the four ways crews actually fail to log: hot-swap shifts, broken meters, missed days, and end-of-week catch-up estimation. Each one has a path that takes less time than guessing. When the easiest action is also the right one, the habit lands.
The companion piece is the manager's weekly review. Without it the data quality drifts within a month, even with a good SOP. With it, the loop closes and the rest of the maintenance program — PM scheduling, cost-per-hour ranking, anomaly detection — has clean inputs to work with.