How to design a fibre sensing trial that supports a decision

A useful fibre sensing trial starts with the decision it must support. Define the operating problem, representative conditions, evidence, responsibilities and acceptance process before testing. Finish with an explicit choice to proceed, revise, obtain specific missing evidence or stop.

Name the decision and its owner

Is the trial deciding whether a proposed approach can address the application, whether existing fibre is suitable, or whether the complete operational workflow is ready? These are different questions. Record the intended decision, who has authority to make it and when it needs to be made.

Define the events and normal conditions

For a perimeter trial, describe both the target events and ordinary site activity. Agree what will be measured, how the location and event will be verified and which conditions the trial will represent. A trial that has not encountered relevant operating conditions should record that limitation.

Specify the observation period and event definitions. Avoid comparing alarm totals from different areas or periods without explaining the differences. Fewer alarms alone do not establish better performance: the trial must also establish what happened to detection of the events the team cares about.

Test the operational chain

Follow the information from detection to operator assessment and response. Check location mapping, interfaces, verification, escalation and responsibility. Identify what should happen if communications, a supporting system or the normal response route is unavailable.

Agree which parts are within the trial scope and which remain assumptions. A successful component demonstration does not automatically validate a complete operating system.

Agree evidence and reporting before the test

Record the test conditions, target-event observations, ordinary activity, system configuration and any changes made during the trial. Agree the evidence needed to support each acceptance criterion and how disputed or incomplete results will be handled.

Do not invent a universal threshold. Required performance depends on the application, the consequences of errors and the customer's operating process. Customer, supplier and relevant specialists should agree the criteria before acceptance testing.

End with a decision record

The final review should state what was demonstrated, what remains unknown and what the evidence supports. Allocate every unresolved item to an owner and a next action. If the proposed approach is unsuitable, identify whether a revised installation, another technology or stopping the project is the appropriate conclusion.

Build the acceptance plan before the test

An acceptance plan should describe what will be tested, what evidence will be retained and who can approve the result. The following is a planning framework, not a prescribed performance standard.

Separate setup from evaluation

Configuration and learning during a trial may be necessary. Record when changes occur and which results were obtained before and after each change. Agree a defined configuration for the evaluation period, with a process for handling essential changes.

For each target-event result, retain the record of what actually happened, who observed it and the system's corresponding output. A screen showing an alert is not enough to assess events that were missed. Record the full set of agreed tests, including unsuccessful and inconclusive results.

What if nuisance alarms fall sharply?

For a hypothetical perimeter trial, an alarm reduction could reflect better discrimination, a quieter observation period, changed settings or lost coverage. The count alone cannot distinguish these explanations.

Check comparable observation periods and conditions, review configuration changes and repeat agreed target-event checks. Assess whether the operator can still identify and act on events that matter. If those checks were outside the trial, record that the evidence is incomplete.

A decision record the project can use

Complete these statements at the review:

• The decision we set out to make was: …
• The configuration and conditions evaluated were: …
• The evidence supports: …
• The evidence does not yet establish: …
• Our decision is to proceed, revise, collect specific missing evidence or stop: …
• Each remaining action has an owner and a decision date: …

Buyer questions

How long should a fibre sensing trial last?

Long enough to address the agreed decision under representative conditions. Duration alone is not proof of adequacy. Explain which conditions were observed, which were not and whether seasonal or intermittent conditions require further evidence.

Can a successful demonstration replace acceptance testing?

A demonstration can establish that a particular event was detected in particular conditions. Acceptance asks whether the agreed project requirements have been met. Use the demonstration as evidence within that wider process, with its limits recorded.

What if a criterion is not met?

Use the agreed process to examine the cause, decide whether a change is justified and define any retest. Keep the original result. Do not quietly move a threshold simply to declare success.

Give the trial a clear purpose

Before equipment arrives, ask Sentari to help structure the decision, evidence plan and responsibilities. Send Robin the application, project stage and decision deadline to discuss the scope.

Discuss your trial plan.

Sources and scope

This is an original advisory framework. It does not certify a system, replace project engineering or set a security acceptance standard.

Related guides: choosing fibre sensing and DAS installed cost.