The problem
Operational data, equipment condition, production context, and maintenance decisions often sit in separate systems. A digital twin must connect them without confusing measured conditions with model-derived conclusions.
Work and evidence
Digital-twin prototypes and simulation-oriented intelligence have been developed, including equipment-state representations. The intended pathway is Measured, then Derived, then Predicted, then Recommended, keeping each category distinct.
Evidence limits
Independent validation, production deployment, and the performance of individual model components are not yet established.
Next validation gate
Document the running prototype, its data sources, model assumptions, state-estimation error, simulation checks, and comparison with observed operations.
Public dataset
The first dataset generated by the flowstation digital twin is now public. Download the Flowstation Twin Dataset v1 (Google Drive). Free access, no sign-in required.
Confidentiality
Asset identities, client data, and proprietary implementation details are intentionally excluded.