A digital twin combines a model of an asset (its geometry, components, physics or business rules) with the live data streaming from it. A twin of a wind turbine, for example, carries the design parameters, the maintenance history and the current vibration and temperature readings, so engineers can simulate load changes or spot drift before a failure. Twins can also be composed: a factory twin aggregates machine twins.
The idea is broader than a device shadow, which only mirrors state. The main cost is keeping the model accurate as the physical asset ages and gets modified, which is a data-quality job as much as a modeling one. Digital twins explained covers use cases and pitfalls.
Part of the IoT Glossary. Spotted an error or a missing term? Tell us.