Smart-Building Pilots Need an Operating Owner

Start with the decision, not the demonstration

A smart-building pilot can look successful while leaving daily operations unchanged. Sensors transmit, a dashboard refreshes and a model produces recommendations. Yet no one has authority to accept the recommendation, arrange access, fund the repair or verify that the problem disappeared. The missing component is not another technology feature. It is an operating owner who is accountable for the complete chain from information to outcome.

JLL’s 2025 analysis identifies business priorities and organisational capability as important to CRE AI programmes [JLL]. That is useful context, not proof that any particular pilot will pay back. Hong and Li’s 2026 communication separately stresses the need to document and evaluate building-AI work [HONG]. Together, these sources support asking what evidence and operating arrangement a local pilot actually needs.

Define one bounded promise

Osmos recommends starting with a sentence that names the decision, user and condition: the engineer will review suspected simultaneous heating and cooling on the selected air-handling units during occupied hours.

This is narrower and more testable than promising an intelligent campus. It identifies which equipment matters, who receives the output and when the output is relevant.

The owner should specify the existing practice before changing it. Perhaps technicians inspect trend logs weekly, perhaps complaints trigger investigation, or perhaps the issue is not currently monitored. Without this baseline, the pilot cannot distinguish a useful new capability from a more attractive presentation of information already available.

Figure 1. The accountable operational loop Original Osmos Global conceptual framework, 2026. Not a measured dataset. Prepared 1 September 2026.

What this means: An alert earns value only through a verified operational outcome.

Separate four responsibilities

The sponsor funds the experiment and resolves cross-functional barriers. The operating owner decides how output enters the workflow. The technical owner maintains integrations and model performance. The assurance reviewer challenges the evidence and checks that reported benefits are not simply supplier claims.

One person may hold more than one role in a small organisation, but the responsibilities should remain explicit.

This arrangement also clarifies what success does not mean. The supplier does not get to define both the metric and its acceptance without challenge. A technician does not become responsible for missing savings when no repair budget exists. IT does not inherit accountability for physical conditions merely because it hosts the platform.

Run a decision rehearsal before connecting

Before live deployment, walk a representative recommendation through the proposed process. Ask who checks it, which information they require, how a work order is raised, what happens if access is unavailable and how the result is confirmed. Rehearse a false alarm and an uncertain recommendation as well as a correct one. The test should reveal dependencies before installation makes them expensive.

Then agree stop, adapt and scale criteria. A pilot may be technically sound but operationally uneconomic because each recommendation requires specialist interpretation. It may need a different workflow rather than a better model. Record these distinctions so the final decision is not reduced to enthusiasm versus resistance.

Illustrative decision rehearsal

Consider an illustrative pilot in which analytics identifies recurring cooling outside scheduled hours. The engineer confirms that the reading is genuine, but the schedule belongs to a tenant team that works irregular shifts. The apparent opportunity cannot be closed by the analytics supplier alone. Workplace, engineering and the tenant must agree what service is required before anyone changes the schedule. This is a decision rehearsal, not a reported case or savings result.

An operating owner would first separate an incorrect schedule from a legitimate business requirement. They would document the service agreement, obtain the required approval and arrange a reversible change only if appropriate. They would then check the resulting conditions and consumption. Without that owner, the same observation can circulate through meetings indefinitely or be acted on without understanding its purpose.

For the first operating cycle, the sponsor should fund time for review and correction, not just data connection. The team should keep a decision log showing who accepted each recommendation, who rejected it and why. Rejections caused by incorrect assumptions are product feedback; rejections caused by missing authority are governance feedback. Combining them as user resistance would hide the actual problem.

At the review meeting, ask whether a competent replacement owner could understand the pilot from its records. If the answer depends on one enthusiastic individual remembering undocumented decisions, the process is not ready for portfolio expansion. Scale should preserve an accountable workflow, not merely replicate the technical installation.

What to report and what remains uncertain

Report recommendations reviewed, confirmed issues, corrective actions completed, persistent outcomes and the effort required. Explain exclusions, unresolved faults and occasions when staff deliberately rejected advice. These measures describe whether the operating chain works; they should not be presented as financial savings until a suitable measurement method exists.

Neither cited source provides a universal pilot duration or success threshold. Equipment criticality, fault frequency, seasonality and data quality determine how much evidence is needed. A short pilot can validate integration and staff usability without validating annual savings. The operating owner’s final responsibility is to state exactly what has been learned—and which claim remains untested.

Source notes

[JLL] Yuehan Wang. Reality check: The true pace and payoffs of AI adoption in corporate real estate. JLL, 2025-10-27. Key highlights; AI pilot selection; Lessons learned. Accessed 1 September 2026. https://www.jll.com/en-hk/insights/global-real-estate-cre-technology-survey [HONG] Tianzhen Hong and Han Li. Good practices for documenting AI-based studies on energy and buildings. Energy & Buildings / Elsevier; author copy hosted by Lawrence Berkeley National Laboratory, 2026-01-20. Sections 2, 3.1–3.6 and 4; pp. 1–4. DOI: 10.1016/j.enbuild.2026.117043. Accessed 1 September 2026. https://eta-publications.lbl.gov/sites/default/files/2026-06/1-s2.0-s0378778826001039-main.pdf

Editorial and visual note

This is original Osmos Global analysis informed by the cited publications. Reported findings are distinguished from Osmos recommendations and illustrative scenarios. Source findings and trademarks remain attributable to their owners. Original visual designs do not imply endorsement by source organisations. The content is general research and does not replace site-specific professional advice.