Timeseries/telemetry data

Data collected by the device is called telemetry. This is data that IIoT devices provide to applications which is collected from the sensors on the devices. Timeseries data is a sequence of data points collected at set time intervals over a continuous period of time. Sensor data is an example of a common way to generate time series data. A time series data store requires a measurement with a corresponding timestamp. The time series service provides an attributes field to include additional relevant details about that specific data point, such as units or site, for example, "Site":"SanFrancisco".

Each source of telemetry results in a tag in the asset model. Telemetry data can be processed at the Edge Gateway, and it is usually sent to the cloud for further processing using various different types of algorithms. Frequency of the data sent varies on the sensor type. Usually it is a data point per second, but this data can add up quickly when we consider the need to store historical data at cloud. We may have to use NoSQL data stores to store a large volume of TimeSeries data.

Time Series data can consist of regular data (data sampled at regular time intervals), or irregular data, for example, data that is recorded only when a certain event occurs (so always at random times).

The Time Series service should provide the following benefits:

• Efficient storage of time series data
• Indexing the data for quick retrieval
• High availability so you can access and query your data from anywhere via HTTP
• Horizontal scalability
• Millisecond data point precision

Typical Timeseries data stores are implemented using the NoSQL DB or using Elasticsearch servers.