Create sensor object types with Ontology Manager

This guidance explains how to use Ontology Manager to create a sensor object type and link it to a root object type. Once you complete the steps below, you will be able to interact with a sensor object type in the platform. In our example, we will create a Flight Sensor object type and link it to a Flight root object type.

Part I: Create the sensor object type

1. Navigate to Ontology Manager, and select Object types from the left side panel.
2. From the upper right corner of the screen, select New object type.
3. In the configuration dialog that appears, configure your object type metadata.
4. In the Properties step, choose to back the object type with the [Example] Sensors dataset that we created with the sensor data pipeline.
5. Then, in the Property field, select the string type unique_sensor_flight_id property to use as the primary key. You can also select Sync all columns from datasource or Map all columns if either of those options is available.
6. Choose the Title column as the title property, which allows the Flight Sensor object types to appear across the application with human-readable names.

7. Depending on your version of the platform you may see options to configure permissions and actions for the Flight Sensor object type. Our example does not require additional permissions or action configuration.

8. Once you have progressed through all the steps in the dialog, select Create.

From the Properties tab of the new [Example] Sensors object type, the properties should appear as in the screenshot below:

Part II: Configure the time series property for the sensor object type

1. Navigate to the [Example] Flight Sensor object type in Ontology Manager and select the Capabilities tab from the left side panel.
2. From the Time series property section select + Add.
3. Select the existing Flight Sensor Series Id property as the time series property, then select Set as default time series property so that it automatically appears in Quiver.

4. Select the time series sync that you created in Pipeline Builder. In our example, it is called [Example] Time Series Sync | Sensor Readings.

5. Select Add property to save the time series property configuration.

Part III: Link the sensor object type to a root object type

Add a link between the [Example] Flight Sensor and [Example] Flight object types using the steps below.

1. From the [Example] Flight Sensor view in Ontology Manager, select the New dropdown menu and choose Link type
2. On the left side of the link, choose the [Example] Flight Sensor object type. On the right, choose the [Example] Flight object type.
3. Set the cardinality for the left Flight Sensor object type as Many, and the right Flight object type as One, meaning the Flight object type has a one to many relationship to the Flight Sensor object type.
4. Set the flight_id column as a foreign key to the Flight Sensor object type, which will set flight_id as the primary key for the Flight object type.

Learn more about link types.

Part IV: Configure the sensor object type

In the time series section, ensure that the sensor object type toggle is on. Set up the Sensor link to use the recently created Flight to Flight Sensor link.

1. Set the link name as the Series Name column. Applications will surface the sensor object data under this series name.

2. Configure units by selecting the Units dropdown menu in the sensor object type setup.

Is categorical and Internal interpolation can be inferred from properties on the sensor object type, but they are not required for this use case. Is categorical is only needed when it is important to delineate categorical time series values from numeric time series values.

Internal interpolation is used to enable applications like Quiver to infer series values between adjacent data points. Review our Quiver documentation on interpolation for more information.

3. Choose to Save the edits to the object type from upper right corner of the screen to view the changes in the Ontology and throughout the platform.

Now, we are ready to use the Flight Sensor and Flight object types in an operational context. Proceed in the documentation to learn how to use sensor object type time series data in Workshop and Quiver.