Azure and the Internet of Things: How Azure Can be Used to...

Azure and the Internet of Things: How Azure Can be Used to Get the Most From IoT

by -
0 190

What do organisations need from Azure to power IoT projects?

Azure and IoT: How Azure can be used to get the most from IoT

Many organisations are looking towards the Internet of Things (IoT) to solve an array of issues and grow business, but does business implement an IoT solution using Azure?

Microsoft has created an IoT platform to serve organisations looking to manage and monitor connected devices. Using the cloud to look after hundreds, thousands, even millions of devices is attractive to businesses which don’t want the complication and cost of scaling such a project in-house. So how does it work and how can we get the most out of it?

Microsoft has made available its Azure IoT Suite since September 2015. There are a number over integrated services for connecting machines and systems, data analysis, and integration with an enterprise’s systems or those of a third party. It is a true platform-as-a-service, allowing organisations the resources necessary to deal with IoT and the data it generates.

Hubs

Arguably, central to the suite is the IoT Hub. This is the interface between IoT devices and Azure. This allows devices to communicate with applications running on Azure. It can scale to millions of devices, as well as monitor and manage them. IT features a device registry and data storage as well as several security features.

Organisations provision an IoT Hub instance, and this can then provision a device. Devices are provided with REST application programming interfaces (APIs) access. There is also the Advanced Message Queuing Protocol (AMQP), for messaging devices as well as MQTT (MQ Telemetry Transport), a lightweight messaging protocol from IBM intended for connections with remote locations where a “small code footprint” is required.

Data crunching

Once data is collected, it has to be processed in order to make sense of it and act upon it. Data from the IoT Hub can be fed into other Azure services.

One such service is Azure Event Hub. This collects, transforms and stores millions of events. It can enable behaviour tracking in mobile apps, traffic information from web farms, telemetry collected from industrial machines, to name a few. It is the front door to an event pipeline that sits between devices that generate events, and applications that consume them.

These events can be piped into Stream Analytics. This is Azure’s real-time event streaming and processing service. It can process and extract information from a data stream to identify patterns, trends, and relationships. These patterns can then be used to trigger other processes or actions, like alerts, automation workflows, feed information to a reporting tool, or store it for later investigation.

Organisations that use Power BI, Microsoft’s cloud-based business intelligence (BI) and analytics platform, can use the service to access interactive dashboards based on the data it produces.

Machine learning

Azure Machine Learning can be combined with Azure IoT Edge to bring intelligence to edge devices. Azure Functions, Azure Stream Analytics, and Azure Machine Learning can all be run on premises via Azure IoT Edge. For example, machine learning can be used to predict machine failure. Data from sensors on an oil rig can be sent to a server on the rig and machine learning models will predict whether equipment is about to break down. Some data can be sent to the cloud to gain an overview of that rig and many others.

Data from IoT devices can be streamed into Microsoft Cognitive Services via APIs to carry out such tasks as image processing and recognition, natural language recognition, text-to-speech, intelligent search and knowledge mapping.

Integration with other applications

Azure IoT suite can also integrate with other systems and applications, such as Salesforce, SAP, Oracle Database and Microsoft Dynamics. This enables businesses to incorporate device data into their businesses processes and workflows.