# Find out where the operating picture breaks. Evaluate OT observability through real investigation questions, missing records, conflicting timelines and the effort needed to reach a defensible next step. Canonical: https://vieox.com/insights/ot-observability/investigation-and-observability-gaps/ Publisher: VIEOX Published: 2026-10-20 Updated: 2026-10-20 ## How can a team identify gaps in OT observability? Replay a real investigation using the information available to its intended users. Record where they cannot establish equipment, time, context, information quality or responsibility. Separate an absent capability from a record that exists but is difficult to find, and prioritise the gaps that change the quality or timeliness of the decision. The most revealing test of an information view may be an investigation that did not go smoothly. Walk through where the team lost the thread: an unavailable period, a changed identifier, an unexplained calculation or an observation that could not be traced to its source. ## Start with a question the team remembers Choose a review with enough history to examine: an equipment concern, a quality finding or an unexpected production shortfall. Ask the people involved to describe what they needed at each stage. Preserve their experience without assuming that a slow answer was caused by an individual’s performance. Use the access available to the intended role. An administrator who can see every system may not experience the same gap as a shift supervisor. The review should establish what the ordinary user can understand within the permissions appropriate to their work. ## Distinguish missing information from difficult access Was the event never recorded, recorded at an unsuitable interval, retained for too short a period or stored under a different identity? Was the information present but difficult to locate? Each description suggests a different next step. Do not turn an unverified search result into proof that a capability does not exist. Record the evidence for the gap and who can confirm it. A specialist may know an existing source that the review team did not recognise. This can reveal useful work already in place and avoid buying or rebuilding a capability unnecessarily. ## Work with the reality of the installed base NIST research on asset condition management identifies integration of older equipment as a challenge and explores differing health-readiness capabilities. That supports examining what a particular asset can reliably contribute before assuming that all equipment will provide the same observations or support the same analysis. Sources: [1](https://www.nist.gov/publications/asset-condition-management-framework-smart-health-ready-manufacturing-systems) ## Ask what would separate plausible explanations A timeline may support several accounts. Write down what additional observation could distinguish them and whether it is proportionate to obtain. A gap is important when it prevents a responsible judgement, not simply because another field could be added to a screen. Hypothetical example: a delayed inspection record could describe work completed late or a timely inspection entered afterwards. The review identifies which time is missing and asks the appropriate owner to clarify it. Until then, the chart should not present either account as certain. ## Evaluate usefulness through the work of the reviewer Observe whether a colleague can locate the relevant period, understand the definitions and explain the limitations without relying on a private conversation. Note repeated copying, manual reconciliation and unanswered ownership questions. These are concrete candidates for improving the information journey. Do not reward a faster answer if it is less supportable. The review should consider the completeness of the account, the remaining uncertainty and whether the next step has an appropriate owner. A confident screen is a poor substitute for a dependable record. ## Turn the findings into a manageable next step Group gaps by the question they obstruct. Identify a responsible person, the proposed correction and how the team will verify that the gap has been resolved. Keep dependencies visible: a new comparison may need a reliable product association before it needs another chart. Return to the original investigation after a change. Can the intended reader now complete the task with the right context? VIEOX welcomes this kind of grounded review because it begins with an operating need and gives any proposed improvement a concrete test. ## Practical starting points - Select a real investigation and its intended user. - Describe exactly where the record becomes incomplete. - Check existing sources before declaring a capability absent. - Prioritise gaps that materially affect the decision. - Replay the question after the agreed correction. ## Questions, answered ### Does a missing dashboard field mean the underlying data is absent? No. The record may exist elsewhere, have another name or be unavailable to that role. Verify the source and access conditions before concluding that the site needs new collection or storage. ### Should every gap be closed before a pilot begins? Set the scope around the question and its consequences. Some gaps can remain visible limitations; others may prevent a useful or responsible evaluation. Agree that distinction with the people accountable for the review. ## Sources and further reading - [NIST — Asset Condition Management: A Framework for Smart, Health-Ready Manufacturing Systems](https://www.nist.gov/publications/asset-condition-management-framework-smart-health-ready-manufacturing-systems) ## Related reading - [OT observability: explore the series](https://vieox.com/insights/ot-observability/) - [Signals events and operating context](https://vieox.com/insights/ot-observability/signals-events-and-operating-context/) - [Industrial AI readiness](https://vieox.com/insights/industrial-ai-readiness/) - [Say hello to VIEOX](https://vieox.com/contact/)