In line with the Cloud First strategy, the Microsoft Azure team is aggressively working towards creating a Connected Systems and Devices platform using the Azure Service Bus, called the Microsoft Azure Intelligent Systems Service. This will enable enterprises to aggregate data from a variety of devices and sensors, and other lines of business applications. The data can then be made as a source for big data analytics using Azure HD Insight.
With all the scalability and performance attributes of Microsoft Azure in play, the platform could address the needs of industries such as transportation, manufacturing, supply chain, retail and healthcare, and enable the creation of smart cities and buildings. The idea of the Internet of Everything is to be able to support more efficient use of resources by providing insight on the data aggregated from them. As Todd Holmquist-Sutherland and Paolo Salvatori included in their Build 2014 conference session, the significance of Internet of Things (IoT) is in building the business of data driven insight. In addition, the significance is not just limited to getting business insights, but also to take appropriate action using the platform. For example, in a smart grid, the system could decide based on insights, the power consumption trends of house-holds and take action to control the ingress.
The Azure IoT stack provides a Cloud Gateway that acts as a secure boundary for communicating with the backend services. The gateway interprets messages from proprietary protocols and sends it to the Ingress Messaging Layer. The Cloud Gateway supports two patterns for communication. A Device Direct pattern where the IP Capable Devices are able to communicate directly with Azure Service Bus for storage and notification, or the brokered Custom Gateway where a set of Cloud Services broker the communication. The gateway allows partitioning of resources that are assigned to specific population of devices, allowing the ability to scale on demand.
The Cloud Gateway interacts with the Azure Service Bus using a topic based publisher subscriber pattern. Messages from the telemetry pump are split into alerts and data, and are sent to the alert processor and storage pre-processor respectively. This video from Build provides a quick preview into what’s in store for the future of IoT with Microsoft Azure.
