Decision and envelope
Define the question, process range, users, outputs, tolerance, update frequency, and excluded uses before connecting data.
Bioprocess digital twin software
Compare the process model with measured data.
Import process records, review their units and map selected values to a saved model. Use supported experimental calibration to compare predictions with separate validation runs. Live plant connectivity and monitoring remain development work.
Process graphDynamic statesImported recordsEquipment basisDecision evidence
Mechanistic state
Connect components, reactions, ODE/PDE states, equipment constraints, schedules, costs, impacts, and uncertainty.
Imported process records
Import authorised historian exports, batch records and assays. Review variable mappings, units, timestamps and source ownership before applying values.
Review the model limits
Check independent validation error, the operating range and unresolved inputs. Record the model version and the evidence needed for the intended decision.
Twin levelAvailable scope and development statusWhat must be proven
Design twinFlowsheet, balances, equipment, schedule, TEA/LCA and scenariosAuthorized design basis, reconciliation and sensitivity coverage
Experimental calibrationSupported parameter fits to declared experimental runs, with separate validation dataResiduals, independent validation data and validity domain
Operational twin · development goalLive sensing, time alignment, monitoring and alerts require a separately implemented site integrationSecure connectivity, data quality, drift handling and qualified response
Closed-loop control · development goalLive equipment control is not an available public Acatian capabilitySite-approved DCS/PLC integration, interlocks, validation and operational ownership
