John Tauchus Digital twin strategy

Start with what the data is for

Most digital twin programs fail at the beginning, not the end: nobody agreed on what the data was for. My approach starts there.

  1. Scope

    Decide what the data must do and who will use it. Not all BIM needs to live forever.

    I classify the effort into one of three tiers:

    Disposable
    Coordination only.
    Transfer
    Handover-ready.
    Operational
    Maintained as a living dataset.

    Then I define the data customer, the system of record, the required fields by asset class, and who owns tagging.

    The gut-check questionIf we deliver an operations-ready dataset, who will use it in the first 90 days, and what decision will it support?

  2. Assess

    Measure where the twin is today on a four-level maturity scale.

    1. L1Spatial model
    2. L2Enriched static model
    3. L3Connected model with on-demand refresh
    4. L4Operational twin: the model becomes the spatial index for live operational data

    The real inflection point is L3 to L4.

    Walk through all four levels in VR

  3. Build

    Move up the path one layer at a time.

    1. Reconcile every document source into a validated foundation.
    2. Strengthen it with reality capture and a targeted, geolocated field audit.
    3. Build the geometric model registered to the scan, with owner parameters and CMMS asset IDs.
    4. Model systems as connected networks.
    5. Close the remaining gaps with a short, specific field visit.
    6. Integrate with BAS, CMMS, and metering.

Score as you go

Each tier gets the scorecard it deserves.

A coordination model is judged on coordination. An operational dataset is judged on normalization, governance, and traceability. Measuring against a known target is what keeps a program from becoming a late-project scramble.

Have a building, a portfolio, or a program that needs its data to agree?

Taking new engagements starting November 2026.

Start a conversation