A timestamp is not proof that two vehicle events align

A graph can place two vehicle signals on the same horizontal axis without establishing that the underlying events occurred together. One timestamp may mark acquisition, another message transmission and another arrival at a recorder. Differences in clocks and buffering can then look like a change in vehicle behaviour.

Before using timing to explain cause or performance, define the observation process. The engineering question is not simply whether every record contains a timestamp. It is whether those timestamps support the comparison being made and what uncertainty remains after alignment.

Identify the event each clock marks

For each source, document when the timestamp is assigned and which clock it uses. Establish whether the recorded time refers to the physical observation or a later stage of processing. A receiving system can timestamp an old observation accurately at receipt while leaving its original timing unknown.

ASAM MDF provides a context for retaining measurement information, while ASAM XIL includes synchronised acquisition capabilities across test-bench ports. Those facilities help teams organise data. Their presence does not remove the need to verify how a specific acquisition setup behaves.

Measure the path between sources

Consider sampling, processing, transport and buffering along the relevant paths. Identify delays that are reasonably constant and those that vary with load or operating conditions. A correction suitable for a constant offset may not address changing latency.

Examine what happens after a restart or loss of synchronisation. A clock can change its relationship to another source while the recorder continues producing apparently continuous data. Preserve events that mark such transitions and distinguish periods where comparison is reliable from those that require caution or a different method.

State uncertainty with the conclusion

Choose timing precision according to the engineering decision. A broad operating trend may tolerate uncertainty that would invalidate a claim about a rapid response. Avoid presenting more decimal places than the observation process can justify. Numeric detail is not a substitute for knowledge of the measurement chain.

When aligning data through an inferred event, record the method and its assumptions. Investigators should be able to distinguish a measured synchronisation relationship from an estimated one. That distinction matters when a later analysis relies on the same data for a different, more demanding question.

Verify the acquisition configuration

Use a suitable reference event or other approved method to check alignment across the sources relevant to the test. Repeat verification when a recorder, network path or software configuration changes materially. The test plan should identify the timing relationships on which its intended conclusions depend.

Keep the verification evidence with the recorded run and analysis. A later reviewer needs to know why two sequences were considered comparable. Time synchronisation discipline protects engineering judgement from a particularly convincing form of error: a clear graph whose visual alignment rests on an unexamined acquisition assumption. The reliable output is a supported relationship between observations, not merely a shared time axis.