A shared acquisition clock does not, by itself, prove that two filtered measurement channels represent the same input event at the same reported time. Ask which filter delays exist, which compensation has already been applied and what each timestamp means. For an integration proposal involving UG35, these are questions for the complete sensor and acquisition chain, not timing guarantees that can be inferred from the aircraft model.

The traces can shift while the clocks agree

Consider a hypothetical bench review in which two channels respond to the same permitted test event. The team confirms that the acquisition devices use a shared timing reference, yet the displayed traces do not line up as expected. One reviewer proposes moving a trace until the peaks coincide. Another proposes replacing the clock. Neither action should be the first response without understanding the signal path.

The useful procurement question is narrower: what evidence supports the alignment claim for this configuration? A statement about the clock answers part of that question. It may not describe what happens to the signal before the reported sample reaches the analysis. Keeping those parts separate makes the supplier's explanation easier to test and harder to overinterpret.

Read delay specifications in their own context

NI's explanation of DSA and SC Express specifications distinguishes analog and digital filter delays, including compensated and uncompensated components. It describes dependencies on configuration, such as gain, frequency or sample rate, for the equipment covered. That documentation illustrates why a single synchronization statement may not describe the complete timing behavior.

It is not evidence that UG35 contains NI hardware, that a particular payload uses those filters or that any named compensation is available in a proposed integration. The review approach below is our editorial application of the distinction. Actual values and behavior must come from the selected equipment documentation and configuration-specific evidence.

Draw the path of one reported sample

Ask the integration team for a simple timing-chain sheet. Begin with the physical quantity at the sensor and end with the record used by the analyst. Between those points, name the sensor, signal conditioning, acquisition device, filtering, timestamp assignment, any compensation and the export stage. The sheet need not expose proprietary internals to identify who owns each boundary.

Use one row per channel when different channels take different paths. A shared clock can appear in both rows while the filter settings differ. Conversely, two channels may use the same processing configuration but differ in how the exported records are labeled. A single box called synchronized system hides both possibilities and gives the buyer little basis for a focused follow-up.

For each named delay or correction, ask whether it is documented, measured, estimated or unknown. These labels are more useful than an unqualified total. They also help prevent a preliminary engineering estimate from being copied into a final acceptance document as though it were a verified system specification.

Record compensation before asking for correction

The most valuable question may be whether someone has already adjusted the timing. Request the exact stage at which compensation is applied and the record that identifies it. If the export is already compensated, a later analyst needs to know that before proposing another shift. If compensation is not applied, the report should say so without implying that a generic adjustment will necessarily solve the problem.

Our editorial recommendation is to maintain separate fields for the original timestamp and any derived timestamp used in analysis. Include the reason, method and responsible process for each transformation. This is a traceability choice for the project, not a claim that all acquisition software offers those fields automatically.

A useful response from a supplier might therefore be conditional: alignment is supported for the specified configuration, with the stated compensation enabled, under the documented test conditions. That is more informative than a broad claim that all channels are synchronized. It identifies what must remain unchanged for the evidence to remain relevant.

Agree a demonstration without prescribing unsafe testing

Ask qualified integration personnel to propose an appropriate, safe test using the actual equipment and its authorized procedures. The buyer's contribution is to define the evidence needed: configuration identifiers, event definition, channel records, expected comparison and acceptance rationale. This article is not an electrical test procedure and does not instruct readers to connect signals, energize aircraft or alter control systems.

A demonstration should preserve the unmodified records as well as the displayed result. Ask which settings were active, what compensation was enabled and whether any additional processing was performed before the plots were produced. Without that record, an attractive overlay may be difficult to reproduce or may depend on a manual adjustment that the production workflow does not include.

Also agree how the team will handle a result that does not meet the proposed criterion. The appropriate next step may be further diagnosis, a revised integration design or a narrower supported use. It should not be an undocumented change made only to make one screenshot appear aligned.

Why aligning peaks is not the whole acceptance test

In an invented review example, an analyst moves one trace until a selected peak coincides with another. The resulting picture looks convincing, but the buyer still needs to know what the adjustment represents. Was it a documented compensation, a fitted offset for that sample or an exploratory visualization? Those descriptions imply different levels of evidence.

Ask whether the method was chosen before evaluating the acceptance sample and whether the same method is part of the intended workflow. Record the reason for the selected event and the limitations of the comparison. A useful review distinguishes demonstration, diagnosis and accepted processing instead of treating them as interchangeable stages.

Do not turn this caution into a blanket rejection of all time adjustment. A documented and justified correction may be an appropriate part of an integration. The issue is whether its purpose and applicability are established. Procurement should ask for the explanation and evidence rather than demanding either an uncorrected plot or a perfectly overlapping one as a matter of appearance.

Keep time scale and reporting statistics separate

Filter-delay review does not settle every timing issue in a dataset. A record can also be misunderstood because its exported time scale is unclear. The companion guide to GPS time and UTC conversion concerns that metadata problem. It should not be used as a substitute for examining the signal-processing path.

Likewise, an aggregate timing report needs clear statistical labels. The discussion of standard deviation versus standard error explains why a smaller bar on a summary plot need not demonstrate smaller variation in the underlying observations. A channel-alignment review should state both the quantity being summarized and the meaning of its reported variability.

For US and New Zealand integration teams, a shared review sheet can reduce ambiguity when equipment selection, analysis and procurement are handled by different organizations. This is coordination guidance, not a jurisdiction-specific approval route. Airworthiness, operating permission and project safety review remain outside the scope of this timing explanation.

Relate the evidence to the proposed UG35 package

The UG35 product page provides the starting aircraft identity for a payload-integration discussion. Do not read the presence of this article as confirmation of a particular sensor interface, data-acquisition device, synchronization output or software feature. Ask the responsible supplier to confirm the proposed configuration and its supporting documentation.

When considering the VTOL and fixed-wing drone collection, compare complete responsibilities as well as aircraft options. Identify who supplies the sensor, who integrates the acquisition chain, who maintains the processing settings and who accepts the data. A missing responsibility can leave the buyer with components that have individual documentation but no agreed system-level evidence.

Keep unknowns visible in the inquiry. If the payload has not yet been chosen, say so. If the receiving analyst requires a particular timestamp convention, include that requirement. This is more useful than requesting a universal synchronization guarantee before the relevant chain exists.

A configuration change should trigger a focused review

Once a configuration is accepted, record which changes could affect the timing claim. The integration owner can identify the relevant filter settings, software versions and equipment substitutions. A later change should prompt a review of the affected evidence, rather than automatic reuse of the earlier acceptance statement.

This need not become an administrative burden. A compact configuration record and a clear owner can make the review proportionate. What matters is that the published claim remains attached to the conditions under which it was supported, and that unresolved differences are not concealed by a familiar model name.

Make the next question specific

Our buyer lesson is to ask whether compensation has already happened before requesting another correction. The strongest inquiry is a two-channel timing-chain description with known settings, timestamp definitions and the intended analytical use. It gives the engineering team something concrete to evaluate without presuming that the aircraft itself solves every timing requirement.

Send that outline through the UNITED UAV contact page and request current configuration details for the proposed UG35 integration. Use sanitized examples rather than confidential operating records. Ask for the unresolved assumptions and the evidence needed to close them, not just a yes-or-no answer to the word synchronized.