THE DIRECT ANSWER
The decision to start with
Accept fleet data by testing the evidence needed for a defined decision: its source, meaning, timing, gaps and export behaviour. Record pass, fail and not tested separately for every vehicle/device group.
Who this is for: Operations leads, dispatch teams and implementation owners validating a fleet platform.
Define the decision and the required evidence
Choose an operational question, such as whether a position is recent enough for dispatch or whether a fuel exception can be reconstructed. List the essential fields and who will use them. Data quality is acceptance against this agreed use case, rather than a promise that every field will exist.
Create one worksheet row per vehicle/device group and required signal. Agree the observation period, review owner and acceptance criterion before seeing the pilot results.
Trace each field to its source
For each required field, distinguish a direct vehicle or sensor measurement from a calculated or inferred value. Record units, conversion rules and the device configuration reference. A visible number is insufficient if its meaning cannot be explained.
Ask the provider to demonstrate the supported source on a representative vehicle. Mark unavailable or conditional coverage explicitly; do not substitute a plausible value for missing data.
Test timestamps and gaps
Keep source timestamps, receipt times where exposed and the export time zone distinct. Reconstruct a normal journey and a parked period using an authorised observation log. Specify the freshness expectation for each case.
Review a safe, approved connectivity interruption if operational procedures permit it. Document which records become available after reconnection and how gaps appear. Do not assume that the last visible value is a new measurement.
Check interpretation across different signals
Ignition state, engine-running evidence and movement can describe different things. Confirm the definition of each displayed state and the actual configured source. A stopped vehicle with an on-state does not by itself prove idling.
Use the Learn Fleet definitions to write test cases for parked, accessory-powered, engine-running and moving conditions where applicable. Have the responsible team approve those observations; record unsupported cases as not tested.
Reconcile an export with the screen
Export one agreed interval through the authorised workflow. Compare vehicle references, field meanings, units, timestamps and a small set of known records with the displayed history. Record any aggregation, formatting or missing fields.
Keep the original export and the comparison note securely. Do not collect unnecessary driver details in a public worksheet. Confirm who can export and what retention or access limitations apply.
Sign off by vehicle group
Summarise passed cases, failed cases and unresolved coverage for each group. A successful location test does not accept fuel readings or diagnostics. Assign a correction owner and retest only the affected cases after a change.
Bring the worksheet to a SafarTrak deployment review. Confirm the signals, configuration and operating workflows supported for the proposed setup before expanding rollout.
Put the criteria side by side
Swipe or scroll sideways to view every column.
| Check | Evidence | Result to record |
|---|---|---|
| Source | Actual signal and configuration | Direct, inferred, unavailable |
| Timing | Known observation and timestamp sequence | Pass, fail, not tested |
| Gaps | Approved interruption and recovery records | Observed behaviour and limitations |
| Export | Same interval on screen and in file | Reconciled fields and open differences |
ILLUSTRATIVE EXAMPLE
A recent receipt with an older position
Illustrative example: dispatch receives a record at 10:15, while its position time is 10:05. The acceptance note retains both values and asks whether the older position meets the agreed dispatch freshness requirement. It does not relabel the position as current. The outcome may differ from an engine-state test on the same vehicle.
Common questions
Is missing data the same as zero?
No. Keep unavailable, missing and measured zero distinct. State which condition the provider reports and how it appears in exports.
Can one vehicle validate the whole fleet?
Only for the coverage actually represented. Different vehicle, device, firmware and sensor groups need their own evidence.
Does this checklist guarantee SafarTrak signal coverage?
No. It is an acceptance method. Confirm actual support for your vehicle, device, configuration and proposed deployment.
Sources and further reading
- Samsara: telematics data types and timestamps
Provider example of separate diagnostic fields and timestamped series; does not establish SafarTrak coverage.
- Teltonika FMC130: ignition and movement sources
Device-specific example showing configurable state sources; confirm the actual tracker configuration.
Published by SafarTrak as general operational guidance. Examples are illustrative; confirm vehicle, device and workflow coverage for your deployment. How we publish our guides.
Download the review sheet (CSV)