Wagering that the factory floor of the future will be buzzing with intelligent, cloud-connected devices, Microsoft is showing off its latest industrial internet of things (IIoT) innovations at Hannover Messe 2018 in Hannover, Germany, April 23-27.
Among the new offerings that the Redmond, Wash., software giant is demonstrating at the industrial technology conference is an automatic discovery service for Azure IoT Suite Connected Factory, said Sam George, partner director of Azure IoT at Microsoft, in an April 23 announcement. Comprising open-source software components and running as microservices on Azure, the service enables manufacturers to quickly and securely incorporate industrial equipment into their IIoT environments. It also provides an OPC UA (Open Platform Communications Unified Architecture) Server interface, enabling support for existing equipment with less manual configuration.
Sometimes, circumstances won't allow for cloud-connected IIoT deployments, whether the quality of a factory's internet connection isn't up to snuff or an organization is restricted in its cloud use by compliance requirements.
Addressing these and other roadblocks faced by manufacturers, Microsoft is making Azure IoT Hub and IoT Hub Device Provisioning Service available on Azure Stack, the company's hybrid-cloud solution. Sold as software and hardware bundles, Azure Stack allows enterprises to run a piece of Microsoft's cloud in their own data centers.
Azure Time Series Insights, Microsoft's data storage, analytics and visualization service for IoT event analysis, is gaining a few new capabilities designed to help businesses makes better sense out of years' worth of device information. The service will soon integrate with Azure Storage, allowing organizations to draw insights from their IIoT equipment data while stashing their data in the cloud in a cost-effective manner.
Borrowing from tiered storage arrangements used in many enterprise IT environments, combined Azure Storage-Time Series Insights implementations will keep data in so-called warm and cold layers, while maintaining access to data for analysis-gathering purposes across both levels. The warm layer is essentially the same data-storage solution used by current customers, said Chandrika Shankarnarayan, principal program manager of Azure IoT at Microsoft, in a separate April 23 announcement.
"We expect most customers to store 30–120 days of data in the warm layer, and 1–20 years in the cold layer, thereby blending the best of both worlds," explained the Microsoft executive. "Cold storage, coupled with device/tag-centric querying across all data, means that customers maximize the cost-benefits of the cloud while still realizing performant querying and trending of historical time series data."
Additionally, Microsoft is working on a device- or tag-centric user experience, enabling equipment operators to zero in on specific devices. Acknowledging that the current interface is geared toward data scientists and analysts, the new interface will enable users to quickly find a desired piece of equipment and view time-series data collected by its sensors.
Microsoft is also hammering away on new integrations that will enable users to perform advanced analytics on their IIoT data.
Azure Time Series Insights will soon store data, based on device and timestamp information, using the popular Apache Parquet historical data file format, Shankarnarayan said. The move will help open the service to other big data services, including Microsoft's own Azure Databricks. For organizations interested in Azure Time Series' predictive-maintenance potential, Microsoft is streamlining the process of integrating the service with Jupyter Notebooks, Azure Machine Learning Studio and other machine-learning solutions, she added.