GPS Device Health Assessment
A tracking system is only as good as the units still reporting, and units fail quietly. This assessment combines how many devices reported today with how complete their position data was, giving you the share of your fleet you can genuinely see, and the queue of vehicles that need a fitter.
Your numbers
Units you believe are fitted, taken from the register. Include the ones on vehicles currently off the road.
Units that sent at least one position in the last 24 hours.
Reporting, but with holes: dropped hours, jumps, or positions that put the vehicle somewhere it was not.
Nothing received for a week or more. Some of these are units that were removed and never written off the register.
What your configured reporting interval should produce over the day. A three-minute interval across 24 hours is 480.
Average actually stored per reporting device per day. Take it from the platform, not from the device spec.
Result
Usable tracking coverage
75.4 %
Reporting share multiplied by data completeness. The share of the fleet you can actually see.
- Devices reporting
- 90.0 %
- Data completeness
- 83.8 %
- Silent devices
- 12 devices
- Intervention queue
- 20 devices
- Missing position reports per day
- 8,424 reports
The formula
- Reporting share = (Devices reporting today ÷ Devices installed) × 100, clamped to 0-100
- Data completeness = (Received position reports ÷ Expected position reports) × 100, clamped to 0-100
- Usable tracking coverage = (Reporting share × Data completeness) ÷ 100
- Silent devices = Devices installed − Devices reporting today, floored at 0
- Intervention queue = Devices with data gaps + Devices silent over 7 days, capped at Devices installed
- Missing position reports per day = (Expected − Received, floored at 0) × Devices reporting today
Assumptions and limits
- Reporting share and completeness are multiplied, so a fleet where every device reports badly and one where most devices report perfectly can land on the same coverage figure. The two components are shown separately for that reason.
- The tool cannot distinguish a dead device from a device on a vehicle parked in a workshop for a fortnight. Some of your silent units are healthy and simply have nothing to say, and only your register knows which.
- Expected reports come from your configured interval, so if the interval was changed and the register was not, completeness reads low for a reason that is administrative rather than physical.
- Position quality is not assessed. A device can report on time, every time, with a drifting or spoofed fix, and this model will call it healthy.
- Gaps and silence are added into the intervention queue without deduplication beyond the fleet-size cap. A unit that was gappy and then went silent may be counted in both inputs.
- Nothing here estimates cost or fitter time. It sizes the problem; how long a truck waits for a replacement unit is a workshop question.
Devices fail quietly, and that is the whole problem
A broken headlight gets reported before the shift ends. A tracking unit that stopped transmitting three weeks ago generates no complaint from anyone, because the vehicle drives perfectly well without it. The failure only surfaces at the moment you need the data: a customer disputing an arrival time, an insurer asking where the vehicle was, a fuel event you cannot place.
This is why device health is a scheduled check rather than an alert you wait for. The fleets that get caught out are not the ones with bad hardware. They are the ones who assumed installed meant working, and never counted.
Silent is easy. Gappy is worse.
A unit that has said nothing for a week is at least honest about it. You can see the gap in a list and send someone. The dangerous category is the device that reports most of the time: it appears on the map, it populates the report, and it drops the two hours that turn out to matter.
Nobody notices, because a partial trip still looks like a trip. Distance totals come out slightly low, which reads as a driver taking a shorter route rather than as missing data. Then a dispute arrives about a specific afternoon and the record has a hole precisely where the argument is. That is why completeness sits inside the coverage calculation rather than in a footnote.
The register is usually the first thing to fix
Run this assessment on most fleets and the silent-device count comes back higher than anyone expected. A meaningful share of those units are not broken. They were removed when a vehicle was sold, transferred to a replacement and never re-linked, or fitted to a trailer that has been parked since spring.
That is worth knowing before anyone is dispatched. Reconcile the device register against the vehicle register first, write off what genuinely left, and re-run. Whatever silence survives that exercise is a hardware queue, and it will be shorter and far more actionable than the first number suggested.
Frequently asked questions
There is no universal number, because it depends on your interval, your cellular coverage and your duty cycle. A fleet crossing long stretches with poor signal will buffer and backfill, so its live completeness looks worse than a city fleet with identical hardware. Track your own figure over time. A coverage number that drops five points in a month is telling you something specific, and no external comparison would have.
In rough order of what fitters actually find: the unit was unplugged, deliberately or during other workshop work; the SIM was suspended or the data plan lapsed; the vehicle sat long enough to flatten the battery and the unit did not recover; the antenna or harness was damaged; and last, the device genuinely failed. The most common cause is not hardware, which is why replacing units without diagnosing the pattern rarely improves the number.
A determined one can, and it is dishonest to claim otherwise. Pulling a fuse or disconnecting the harness stops transmission, and a jammer suppresses it with no physical trace. What is much harder is doing it invisibly: the gap has a start, an end, and a vehicle that arrives somewhere it could not have reached without moving. Tamper detection and battery-backed reporting narrow the window rather than closing it, and any vendor promising a device that cannot be defeated is describing a product that does not exist.
Weekly for the silent list, monthly for completeness. The silent list is short and actionable, so it belongs with whoever runs the workshop diary. Completeness moves slowly and rewards a trend view. Reviewing either quarterly means a unit can be dead for most of a quarter, which is exactly the span over which the data you are missing turns out to have been needed.
GPS Tracking & Telematics keeps a device-health view: last-contact time per unit, reporting completeness, and units flagged as silent beyond your threshold. What it cannot do is decide whether a silent unit is a broken device or a vehicle that has not moved since March. That still needs someone to reconcile the device register against reality. The platform can only tell you which rows to look at.
Stop estimating. Measure it.
GPS Device Health Assessment gives you the arithmetic. GPS Fleet Tracking Software gives you the live numbers from your own fleet.