Diagnostic
Is the twin a live system or a presentation layer?
Score the operating disciplines behind the model, each from 0 (absent) to 5 (owned and consistent), then enter the timings of the pipeline. The point is not visual quality. The point is whether site state reaches the decision workflow in time to matter. The lens follows the information-management discipline of ISO 19650. Latency is computed from the timings and gates the result, so a slow twin cannot rate as a live system however strong the other layers look.
Pipeline timing iThree timings set how old the information is when someone acts on it. The median time from a field event to the model is the processing time plus half the capture interval. That number gates the result: a slow twin cannot rate as a live system.
Current state
Pilot System
The model has useful ingredients, but the workflow still depends on manual handoffs and late updates.
Age of the information iThe model shows the latest capture that has finished processing, so its age climbs until the next one lands and then drops. A field event waits for the next capture, then for processing, then for the next decision meeting. The median and worst-case waits follow from the three timings.
Model ageDecision meetings and the 24 and 72 hour lines
Meeting days on the chart are an example. The figures assume meetings fall anywhere in the capture cycle, and that a work area changes state at random about once per change interval.
Layer health
Next controls to write down
Foundations
The layers and the latency gate in this diagnostic draw on:
- ISO 19650: information management for BIM and digital delivery
- Four-layer twin architecture: capture, process, integrate, act
- UAV pose-estimation & computer-vision progress capture: Lan and Awolusi (2025), CIB World Building Congress; Lan and Sulbaran (2023), CONVR
- Common Data Environment (CDE): single source of project truth
- Age of Information: Kaul, Yates, and Gruteser (2012); Yates and others (2021). The median and worst-case ages here follow from that work for a fixed capture interval and a fixed delay
Diagnostic use only. Scores are directional and should be calibrated against the project's actual schedule, RFI log, capture plan, and decision cadence.