Keep the evidence with the next decision.

Smart-building adoption and actionable signals

Adoption is observable in everyday work. Take one signal and ask whether a technician can find the correct building and asset, understand the observation, and reach an assigned next step. A dashboard viewed regularly is useful activity evidence; a verified operational response is a different outcome.

For a hypothetical out-of-hours air handler, check point identity, timestamp, units and coverage before investigating the schedule. Record occupancy changes and sensor uncertainty. An unexpected value may reflect a mapping error or an approved operating change.

Proposed review: follow five recent observations through your existing process. Count how many reached an owner, how many were duplicates, and how many had a documented follow-up. Keep unresolved cases visible.

AI explanation, operator oversight and continuity

A recommendation should expose the evidence behind it: observations, asset mapping, relevant history and missing information. Distinguish a measured condition from a possible cause. A plausible explanation can still be wrong. NIST AI RMF provides a general risk-management reference; it does not certify a building product.

Define who may accept, challenge or reject a proposed action. Preserve the operator correction and its reasoning. During a vendor or platform change, test whether asset identifiers, units, timestamps, access rights and history remain usable. Data transport alone does not establish interoperability.

Open reflection: which information would make your operator decline a recommendation? Use those cases as evaluation material before expanding an AI pilot.

Maintenance effectiveness, failure modes and quality

PM completion tells you that an activity was recorded. Compare it with inspection findings, recurring defects, unplanned repairs and equipment behaviour to assess whether the activity addresses the intended problem. Keep the distinction between a good task performed inconsistently and a task that does not address the failure mode.

For each selected asset, connect the task to the failure mode it is intended to detect or prevent. Record findings and later failures against the same equipment. Review the task, interval and execution with qualified personnel and the applicable equipment instructions. Site-specific conditions can justify documented adjustments to a shared portfolio procedure.

Proposed review: choose one recurring task and examine the last several records. Did the findings change the next action? Did the same defect recur? Scheduled, condition-based and corrective approaches should be chosen by consequence, evidence and constraints rather than by a single universal rule. Predictive maintenance remains a research area for Logic Foundry.

From inspections to owned repairs and critical equipment history

An inspection finding needs an asset, condition, priority, responsible person and next action. When repair requires parts or an outage, preserve the open finding and the plan. Closing the inspection should not make an unresolved defect disappear from the team review.

Keep service evidence with the equipment: component changes, test results, inspection findings and outstanding actions. A new technician or service provider should be able to understand what changed and what still needs attention. Share records only with appropriate access and review documents before external distribution.

Proposed review: trace one critical asset from its last finding to the repair and the verification evidence. Separate completion of the activity from confirmation that the condition was resolved.

Critical spares and obsolescence

A serviceable asset can become vulnerable when an essential replacement component is unavailable. Review critical equipment with parts references, supplier lead times, alternatives and replacement planning. Keep the operational consequence of downtime visible alongside purchase cost.

Compatibility needs technical review. A similar part name or appearance is not enough to approve a substitute. Identify who can assess compatibility and what documentation they require before a failure creates urgency.

A useful maintenance-strategy discussion starts with the consequence of failure, detection opportunities, repair access and parts availability. This is a planning framework, not a claim that CORE-FM includes a spares optimisation system.

Performance baselines, benchmarks and portfolio sustainability

A portfolio comparison needs context: building use, occupancy, weather, operating hours, meter boundaries and data coverage. Identify whether the measure is energy, peak demand, emissions or cost. These quantities answer different questions and should not be treated as interchangeable.

Connect each improvement to its building, affected equipment, owner and review date. Record the baseline period and changes that could explain the result. A completed project or a certification provides a reference point; continued performance requires monitoring and shared operating responsibility.

Proposed review: inspect one performance claim and reconstruct the source, measurement period, coverage gaps and comparison method. The Berkeley Lab campaign findings are useful evidence about participating U.S. programmes, not savings promises for your portfolio or CORE-FM.

Retrofit and energy-project handover, demand reduction and cold climates

A useful handover includes updated asset records, maintenance instructions, control documentation, commissioning evidence, operator training and a named review owner. Make the intended operating schedule and comfort expectations explicit, including responsibilities shared by owners, occupants and service providers.

Demand reduction needs follow-up because occupancy, schedules and equipment condition can change. Compare seasonal operating evidence with the assumptions behind the project. In cold climates, include winter conditions, heating requirements and operator feedback in the review plan. A short warm-weather review cannot establish winter performance.

Proposed handover exercise: ask the operating team to reconstruct one design decision, its constraints and the evidence needed for a seasonal check. Any control adjustment belongs with qualified personnel and existing authorisation, ventilation, comfort and equipment requirements. DOE re-tuning guidance offers a federal-building process reference, not universal engineering instructions.

Data provenance, coverage and future models

A report should carry source identity, timestamp, units, asset context, processing assumptions and incomplete coverage. Explain gaps rather than quietly substituting assumptions. Reports sent beyond the facilities team need enough context for the next reader to judge what the evidence supports.

Start a digital-twin proposal with one decision, then define model scope, update frequency and acceptable uncertainty. The NIST manufacturing workshops identify useful questions about interoperability and validation; applying them to facilities management is our inference. A realistic display is not evidence that a model is current or fit for purpose.

For a proposed BMS integration, separate read-only observation from control authority. Test stale sensors, incorrect mapping and loss of connectivity. Current CORE-FM capabilities are properties, assets, work orders, preventive maintenance and sensor monitoring. Confirmed research spans predictive maintenance, AI-assisted work orders, anomaly detection, energy optimisation, digital twins, BMS API integrations and future smart controls, reporting and decision tools. These areas do not establish shipped integrations, autonomous controls, an open customer API, measured savings or release dates.

Continue the investigation

Read our report on the future of facilities management for the evidence, limitations and a proposed small observational evaluation.

CORE-FM task guides cover current application workflows. This field guide addresses the operating questions around those tasks.

Primary references

• [1] IEA (2025), Energy Efficiency 2025: Buildings. Global energy context, 2024 consumption.
• [2] Berkeley Lab (2020), Proving the Business Case for Building Analytics. U.S. Smart Energy Analytics Campaign, 2016–2020.
• [3] U.S. DOE FEMP, Re-tuning in Federal Buildings. Operational guidance, accessed October 8, 2026.
• [4] NIST (2023), Artificial Intelligence Risk Management Framework (AI RMF 1.0).
• [5] Frechette, Shao and Pease, NIST (July 21, 2026), Digital Twins Workshops Summary Report, NIST IR 8620. Manufacturing research.

References reviewed October 8, 2026. External authors and organisations are sources, not customer endorsements.