Digital Twin Build Reference

This section is the detailed reference for building a Digital Twin — the in-depth companion to Step 5: Build Digital Twin. It follows the order you actually work in: choose the data, understand the kind of twin it produces, set the configuration, and tune it for a good fit.

Everything here happens in the twin builder, a four-step wizard (how to build, choose your data, add domain knowledge, review and run). Two paths lead into it for a new twin: New Digital Twin on the Digital Twins list, or Create & build twin from the datasets the Ontology page suggests under Datasets you can make. A third, Modify model on an existing twin, reopens the same builder seeded with the current configuration and creates a new version when run — except on an untrained 1.0.0, which is replaced in place rather than versioned up.

• Selecting your Dataset — the one choice that triggers everything, and what the platform auto-detects when it reads the file.

• Types of Digital Twins — Static, Temporal, and the two Multi-Environment variants, and how the platform decides which fits your data.

• Configuration to Build Digital Twin — every control on the build screen: environments, fields, training options, algorithm versions, and variable roles.

• Tuning for Fit — the change-config, retrain, read-fit loop, and the levers that move the result most.

Once a twin is built, see Configuration for Existing Twin and the other Digital Twin pages for working with it.