Digital twin builder (preview) tutorial part 4: Explore your ontology

Now that you have an ontology with entity types and relationship types, you explore it in this tutorial section.

Use the Explore mode

Select Explore entity instances from the menu ribbon to switch to the Explore mode. The view for this mode contains a search bar in the top right corner for entity type and keyword searches, and an Advanced query experience.

First, filter the assets by entity type.

1. Choose Distiller from the dropdown menu.

2. Observe that the result set contains all 10 instances of the Distiller entity type.

   

Next, search on display ID, which is a concatenation of the source columns that you chose during mapping.

1. With Distiller still selected from the dropdown menu, enter D101 into the search bar.

2. Observe that the result set contains one entity instance with that name (distiller D101).

   

Next, try an advanced query, which lets you filter on multiple entity type properties.

1. Select the Advanced query button.

2. Select the Condenser entity type from the dropdown menu.

3. Add a filter for a cooling medium that's water-based. Under Filters, select Cooling medium as the property, leave is as the operator, and enter Water as the value.

4. Add another filter for the installation date. Select Add filter, which adds a new filter row automatically linked to the previous filter with the and operator. Select InstallationDate as the property, starts with as the operator, and 6/2/2018 as the value.

5. Select Apply to run your query.

6. Observe that the result only shows one condenser, which matches your filter criteria.

   

Keep this view open for the next step.

View entity instance details and time series data

You can select an entity instance to view its details and any associated time series. As a result of your mappings so far in this tutorial, these entity types contain modeled properties and data values.

First, view the details of an entity instance.

1. Select the condenser result card from your most recent query to open the Details view.

2. Observe that this view shows entity instance properties, like the site ID and cooling medium.

   

3. Keep this view open for the next step.

View related time series charts.

1. Using the tabs on the top left corner of the page, switch from Details to Charts.

2. In the date range filters in the top right corner of the Charts tab, change the Start date time to 01/01/2025 12:00 AM and the End date time to 01/12/2025 12:00 AM.

3. Check the box next to the Pressure time series name to view its data in a time series chart.

   

   Tip

   If you don't see any time series data, rerun the Condenser time series mapping. You can check its status in the Manage operations page, accessible from the semantic canvas ribbon.

4. Select the Pressure and Temperature time series names together and visualize multiple data sets on the chart at the same time.

5. Experiment with changing the date range and toggling between different aggregation functions like average, min, and max.

Deactivate entity type

Sometimes an entity type is misconfigured or no longer needed. In that case, entity types can be deactivated. Deactivating an entity type removes the entity type and its mapping configurations from the semantic canvas. It also removes entity instances and time series from the explorer.

Enact that scenario now by removing the Reboiler entity type from your ontology.

1. Select Home in the top left corner to return to the semantic canvas. Select the Reboiler entity type to view it in the canvas.

   

2. Before you can deactivate an entity, you must delete its relationship types. Select the isPartOf relationship type in the semantic canvas to bring up the Relationship configuration pane, and select Delete relationship.

   

3. After deleting the relationship type, you can deactivate the entity type. Hover over the Reboiler entity type and select the ellipses next to it. Select Deactivate entity from the menu, and confirm when prompted.

   

4. The entity type disappears from the Entities list and is no longer visible in the semantic canvas.

   

5. Switch to the Explore mode. Observe that you can no longer filter to the Reboiler entity type or search for them by name.

View underlying data

The ontology data for a digital twin builder (preview) item is stored in a Fabric lakehouse associated with the digital twin builder item. In that lakehouse, you can view your ontology data through the domain layer, which is a set of views that directly reflect the logical structure and relationship types defined in the domain ontology. For more information about digital twin builder data storage, see Modeling data in digital twin builder (preview) - Storage and access.

Follow these steps to view your domain data.

1. Go to your Fabric workspace and identify the SQL endpoint of the lakehouse associated with your digital twin builder item. The SQL endpoint has the same name as your digital twin builder item with a dtdm extension, so in this case it's called TutorialDTBdtdm.

   

2. Select the SQL endpoint to go to the explorer page.

   In the navigation pane under Schemas, the dbo entry represents the base layer and the dom entry represents the domain layer. You need to use the dom layer to simplify creating Power BI reports, so expand dom.

   Underneath dom, open the Views section. Observe that each entity type in your digital twin builder is reflected as two views: entityname_property and entityname_timeseries. The relationships view captures all relationship instances.

   Within the rows of each entity type view, there are values for EntityInstanceId1 and EntityInstanceId2. Together, these values form the unique ID of each entity instance.

   

3. Explore the views using the UI (or SQL queries) to check that the sample data is mapped correctly. Verify the following items:

   1. Your mapped properties exist inside the views for their associated entity types, and have data mapped to them (except for fields that were left unmapped).
   2. Entity instances have no null values for the EntityInstanceId1 and EntityInstanceId2 columns.
   3. The relationship view contains relationship instance data, and there are no null values. Notice that each row references a relationship instance between a source entity instance (identified by the combination of FirstEntityInstanceId1 and FirstEntityInstanceId2) and a target entity instance (identified by the combination of SecondEntityInstanceId1 and SecondEntityInstanceId2).
   4. There's no view for Reboiler_property or Reboiler_timeseries, since you deactivated the Reboiler entity type.

Now that you've explored the structure of the digital twin builder data, you're ready to use that data in a Power BI report.

Next step