CONNECTED TECHNOLOGY. CLEARLY EXPLAINED.
Evaluate an industrial IoT pilot
Prepare a decision brief through six steps: problem and baseline, architecture, operational approval, measurement, ownership and a fair go/no-go review.
Learning path · Business and industrial · By Admin · Reviewed 20 October 2026.
Start with the operating decision the pilot should improve. Work through the six readings and build one brief that a process owner, engineer and buyer can discuss together. The exercises prepare an evaluation; they do not substitute for a site survey or establish a measured return. The path is self-directed and issues no credential.
Before you start: Familiarity with the process or access to a responsible process owner. These written exercises can use a hypothetical, non-confidential example. Site work needs the organisation’s own qualified people and approved procedures.
What you will make: A pilot decision brief with a bounded problem, baseline, system/responsibility map, approval questions, measurement plan, operating owner and explicit decision criteria.
Make the pilot answer a decision
Agree the baseline
Describe the present process, review workload and relevant operating conditions. Decide which observation would change a decision. Without a comparable starting point, an attractive chart cannot establish a useful improvement.
Map the complete responsibility chain
Identify sensing, connection, processing, application and the person allowed to respond. Include the existing operating process, qualified engineering support and continuing maintenance. A manufacturer, seller, installer and service operator may have different roles.
Define the exit as well as the trial
Set a review point and written go/no-go criteria before the pilot begins. Include evidence quality, useful completed actions, burden, support and full continuing cost. Preserve a practical route to stop, recover or export the agreed records.
Six steps, with something to show at each stage
1 · Bound the problem and baseline
Read the industrial-pilot article and choose one asset group, process question and accountable decision owner. Document the present review process and operating conditions. Keep proposed benefits as hypotheses until comparable observations support them.
Do: Write the decision, baseline, observation period, useful response and exclusions on one page.
Checkpoint: The process owner can explain how a result will affect a permitted operating decision.
2 · Draw the architecture and boundaries
Use the gateway guide to draw the measurement, device interfaces, communication, processing, application and work record. Label the owners and relevant IT/OT boundaries. Identify what stays local, what leaves the site and what needs a separate integration agreement.
Do: Create a system/responsibility map and list unknown interfaces, access paths and data ownership.
Checkpoint: Every handover has a named owner and a question that can be checked against the actual proposed configuration.
3 · Plan operational approval
Use the OT-security guide to prepare questions for the responsible operations, engineering and security teams. NIST’s OT guidance explicitly considers performance, reliability and safety alongside security. Keep site access, installation, connectivity and change approval within the organisation’s established process.
Do: Record the necessary reviewers, change boundary, approved observation approach and recovery responsibility.
Checkpoint: The pilot has an agreed approval route; no proposed monitoring feature silently becomes a control instruction.
4 · Design evidence that can be reviewed
Use the fan-monitoring pilot as a worked planning example. Specify source identity, time, units, operating context, data-quality states and the allowed human response. Define how missing or unsuitable records and false exceptions will be investigated.
Do: Write a test/evidence table with the setup, expected observation, reviewer and record to retain. Leave actual results unfilled until observed.
Checkpoint: A reported alert can be traced to its source and context, and useful completed review is distinct from alert count.
5 · Assign ongoing operation and cost
Use the manufacturing guide to include the people and processes after installation: configuration ownership, review workload, fault handling, replacements, updates, support, export and exit. Ask the responsible provider to state the exact delivered scope and exclusions.
Do: Create a responsibility and cost table covering setup, connectivity, software, support, maintenance and end-of-service handling.
Checkpoint: The brief names who keeps the system useful and what effort or cost remains when the initial trial ends.
6 · Make the go/no-go review fair
Use the evidence framework to separate vendor documentation, proposed design and actual observations. Compare the recorded baseline with the agreed pilot conditions, include false/stale exceptions and review effort, and explain any limitations. Keep absent evidence explicitly unassessed.
Do: Write the decision criteria, who approves expansion, what would trigger another test and what would justify stopping.
Checkpoint: The decision can be traced to evidence and responsibilities rather than a feature count or an unsupported saving claim.
Read the finished brief as a team
Bring the baseline, system map, approval questions, evidence table, responsibility/cost table and decision criteria together. Check that the trial can answer the original question within the approved operating boundary.
Keep the standards context current
The linked NIST SP 800-82 Rev. 3 publication is final guidance from September 2023. Its current publication page points to an initial public draft of Revision 4; a draft is not the final adopted baseline. Ask the responsible organisation to choose its applicable current standards and procedures.
Request a scope that can be checked
Use a non-confidential requirements brief when seeking help. A proposal should identify the exact equipment and services, acceptance record, delivery responsibilities and support/exit terms. A directory profile or an enquiry does not establish a supplier appointment.
Sources and scope
The exercises and decision brief are original editorial work. The NIST OT publication overview supports the need to consider performance, reliability and safety; IR 8259A supplies device-capability context. Linked IoTPapa guides connect the stages. No plant data, site assessment, installation, operational trial, security validation, measured financial return or learner credential is reported.
NIST SP 800-82 Rev. 3: OT security publication and revision status
Keep the decision moving
NEXT READING
Work through the fan pilot
Follow observations into an accountable maintenance review.
NEXT READING
Explore another industry
Apply the decision framework to a different operating context.
NEXT READING
Prepare a private enquiry
Describe the broad problem and evidence gap for Admin review.