The Digital Twin Is Not the Authority

The physical building exists whether or not a digital model is operating. Its equipment, occupants, atmospheric conditions, and operating history are the reality that engineering systems attempt to understand. A sensor can observe a condition, but the instrument reading is not the building itself. A model can approximate the building, but its simulation is not physical reality. We begin here because every subsequent claim must be traceable to what actually existed. We must distinguish observed conditions from inferred conditions, and preserve the original state before a consequential intervention. The governing principle is simple: reality precedes representation. TA14 does not need to own the building to examine whether a proposed consequence is sufficiently supported.

An Atmospheric Integrity Record, or AIR, is an environmental evidence governance construct, not a new sensor or dashboard. Existing instruments observe temperature, carbon dioxide, particulate matter, pressure, and other conditions. AIR establishes requirements for preserving observations with instrument identity, timestamps, operating context, continuity, uncertainty, and relevant changes. Gaps must remain visible. Raw observations and adjusted values must remain distinguishable. An AIR does not make every reading accurate or admissible. Instead, it makes the evidence and its limitations examinable. Digital twins and analytics may consume governed records without TA14 becoming their database or application. The distinction is preservation and examination of evidence, not ownership of the technology that measured it.

A digital twin may combine models, measurements, simulations, and predictions to represent a physical facility. It can discover patterns and propose changes to ventilation or other systems. That is a valuable engineering capability. But a prediction is not an observation, and a recommendation is not authorization. The model may rely on incomplete inputs, uncertain assumptions, or changed conditions. TA14 therefore preserves the boundary between what the twin represents, what the underlying evidence establishes, and what a responsible party is permitted to execute. Understanding is not permission. Capability is not authority. A digital twin remains an independent tool, while governance examines the proposed consequence.

When a digital twin proposes increasing airflow, the suggestion must first become a bounded proposed consequence. Identify the asset, the precise requested action, the scope, timing, evidence basis, requesting party, and operating constraints. An attractive prediction or apparent benefit does not grant permission. The proposal is then eligible for an independent examination of admissible evidence, applicable authority, and established standing. A recommendation may be held, denied, escalated, or allowed within specific bounds. None of those dispositions is created simply by the intelligence of the model. TA14 governs the crossing from a technical recommendation toward a real-world consequence.

The TA14 governing question is: Does this proposed consequence have sufficient Admissible Evidence, Applicable Authority, and Established Standing to become reality NOW? Evidence must be sufficient for the specific purpose, not merely available. Authority must apply to the actual intervention and current context. Standing must establish who is entitled to request or commit the action. The examination may result in ALLOW, HOLD, DENY, or ESCALATE. These are bounded dispositions, not generic certifications. A prior permission does not necessarily survive changed circumstances. The authority boundary is independent of the digital twin, the sensor, and the controller. No admissible evidence. No admissible execution.

Authority context can cross organizational and technical boundaries, but execution authority must be established locally. A researcher may supply observations. An engineer may develop a simulation. A digital twin may recommend a change. An external governance examination may assess the proposed consequence. Yet the actual building system and responsible operator must establish the authority and constraints for that local action. AEA does not replace the building automation system, controller, actuator, or human decision maker. It governs the conditions under which the consequence may proceed. An execution receipt records an action, but it does not prove that the intended result occurred.

After a locally authorized intervention, we must observe the actual physical response. An execution log may establish that a command was issued or a setting changed. It does not prove that indoor air improved, that pressure constraints remained satisfied, or that unintended consequences were avoided. Post-execution observations need their own preserved evidence, chronology, context, and uncertainty. Compare the resulting conditions with the original baseline and the stated acceptance criteria. The outcome may be improved, restored, unchanged, degraded, or unresolved. An outcome cannot be inferred solely from the command history or the model prediction. Verification returns us to new physical reality, which becomes the basis for any future proposal.

The complete architecture separates responsibilities rather than merging technologies. Sensors observe. Environmental records preserve what can be established. Digital twins represent and simulate. Proposed consequences are defined. AEA examines evidence, authority, and standing. Existing locally authorized systems execute. New records support outcome verification. Across this process the TA14 governing chain remains Reality, Record, Continuity, Admissibility, Binding, Commit, Execution, Outcome. The overlay can potentially apply to hospitals, schools, factories, public infrastructure, and other domains without TA14 becoming their sensor, artificial intelligence, digital twin, or controller. This showroom is an independent educational architecture, not proof of an operational integration or institutional endorsement.