Time Series Data Stream
Definition
An important feature of NuvIoT, is the ability to collect data and then use it to improve performance. Time-Series Data Streams are an optimized storage technology that is very powerful for understanding the behavior of IoT Devices over a period of time. Every data point recorded in this type of Data Stream must have a time stamp. These data points may than be grouped together for analysis. This sampling can be thought of as a Time Bucket which holds all the recorded data within a predefined period of time. Operations may be performed on the Time Bucket, such as averaging, minimum and maximum values or standard deviations.
For more in-depth information, click here: https://docs.microsoft.com/en-us/azure/architecture/data-guide/scenarios/time-series
Example
Vibrations produced by a Plymouth IoT Conveyor Belt were recorded in a Time Series Data Stream. Machine Learning analyzed the Time Series Data Stream to determine an acceptable level of vibration. When the Time Series Data Streams began producing consistently higher levels of vibration, an Error Code was produced.