Model a physical environment as a digital twin in Azure Digital Twins using DTDL and query the graph

Steps

1. Author DTDL (Digital Twins Definition Language) model files in JSON-LD defining interfaces with properties, telemetry, relationships, and components; use DTDL v2 or v3 and validate the files with the DTDL parser SDK before uploading
2. Upload models to Azure Digital Twins via the CLI (az dt model create --dt-name <name> --models <file.json>) or REST API PUT https://<host>/models; models are immutable once created — decommission old versions and create new ones for changes
3. Create twin instances corresponding to real physical entities: az dt twin create --dt-name <name> --dtmi dtmi:example:Room;1 --twin-id room1; set initial property values with --properties '{"temperature":21.0}'
4. Create relationships between twins to model spatial or logical topology: az dt twin relationship create --dt-name <name> --twin-id room1 --relationship-id rel1 --relationship contains --target sensor1; relationships mirror the DTDL relationship definitions
5. Query the twin graph using the Azure Digital Twins Query Language (SQL-like syntax): SELECT * FROM digitaltwins WHERE IS_OF_MODEL('dtmi:example:Room;1') returns all room twins; JOIN traversal up to 5 hops for relationship path queries
6. Integrate live telemetry by routing IoT Hub messages to a function that calls the Update Twin API to patch telemetry-backed properties; use Event Routes in Azure Digital Twins to forward twin change notifications to Event Hub or downstream services