Internet of Things (IoT) analytics is the practice of gathering data from connected devices and analysing that data to gather insights or aid decision-making.
Typically associated with industrial IoT, IoT analytics involves placing sensors on connected devices, which makes it possible to collect data from manufacturing systems, vehicles, oil rigs, cameras, pipelines, weather stations, smart meters and even mobile phones and smartwatches.
Once collected, this data can be combined with more typical sources of data such as database systems and sales transactions, or it can be analysed independently.
IoT analytics grabs data from edge devices that are usually “connected”, which has led to the ability to monitor and analyse mobile systems and remote sites in real time.
Previously, connected device data would be collected and stored in a repository like HDFS for subsequent analysis. How technology has progressed! For example, Apache NiFi now allows for IoT data to be analysed in real time while it is being streamed.
This, in turn, means that IoT analytics can be used to make predictions and insights to aid decisions that can affect live processes. As an example, predictive IoT analytics may be able to identify an impending failure in a manufacturing process before it happens, allowing companies to mitigate, keep manufacturing processes functioning while navigating and implementing preventative maintenance measures before failure occurs.
In the case of airlines, taking sensor data from planes in flight may even lead to preventive steps being taking during flights before a dangerous situation occurs.
The key with IoT analytics is to have an infrastructure that can connect to, manage the streams and analyse data from thousands of devices in real time. Hortonwork’s DataFlow is based on Apache NiFi and facilitates this requirement. They are using this technology to enable IoT and data in motion analytics for numerous companies in real-world use cases.
You can learn about some of these examples here.