Unified Namespace: Driving operational excellence from the factory floor to the boardroom

The Opportunity

In today's fast-paced digital landscape, organizations must embrace change and adopt innovative strategies to remain competitive and thrive. Companies with significant physical operations like manufacturing, retail, and logistics have spent the last several years investing in instrumenting those operations and connecting them to the cloud. This has allowed organizations to collect and act on large amounts of data generated within their physical operations.

While this has provided value inside of projects that target very specific improvements, such as providing better maintenance for complex machines, it has proven extremely difficult to create value for opportunities that stretch across multiple sites or disciplines. One of the primary reasons that has been difficult is that the data, coming from multiple sites, different processes, and roles, is rarely well organized and usually lacks context that can be easily understood outside of the local environment.

Consider a large automotive manufacturer with multiple plants, each using different systems and protocols, such as programmable logic controllers (PLCs), supervisory control and data acquisition (SCADA), and manufacturing execution systems (MES), for data management. In this scenario, it is not uncommon for engineers to gather data manually from each individual system through inefficient point-to-point integrations, leading to delays in accessing the data needed for timely decision-making and issue-resolution. This is a time-consuming process, but necessary when systems operate in silos, using disparate protocols, and lacking real-time data accessibility.

The Solution: Unified Namespace

To address this challenge, the industry has proposed a framework called Unified Namespace (UNS) that provides a common way of understanding data across disparate systems. UNS is emerging as an industry framework for fostering a culture of collaboration and continuous improvement that is vital to driving digital transformation within an organization, unlocking the full potential of the data held across a complex enterprise.

Organizations can implement UNS to modernize their operations and easily leverage data generated across multiple locations, processes, and roles. By providing a centralized, standard approach to data management, UNS breaks down data silos, enables real-time data access across the entire organization, and simplifies integration between diverse systems. This not only reduces integration costs and technical debt, but also empowers organizations to make data-driven decisions, optimize processes, and quickly respond to changes.

Benefits of Unified Namespace

Back to our example of a large automotive manufacturer, after implementing UNS, all data from the various systems is organized and available in a common way. Now, when an issue (such as within production) occurs, their engineers can quickly access data from across the entire organization, regardless of the original source or protocol. This enables rapid diagnosis for problem solving, minimizing downtime, and improving overall equipment effectiveness (OEE). Without UNS, when a production issue occurs, those same engineers would have to extract those data points from specific databases or machines on a case-by-case basis, adding to cost and delaying resolution.

Additionally, UNS facilitates seamless integration of new IoT devices and sensors through a defined data schema, enabling efficient incorporation into the enterprise data estate. This scalability supports continuous monitoring, process optimization, increased efficiency, and cost savings. This unified data access fosters collaboration between departments. For example, the maintenance team can now easily share equipment performance data with the quality control department, enabling them to identify potential issues before they impact product quality.

Principles for adoption of Unified Namespace across industries

While UNS is not yet a standard, it is an evolving pattern that many customers in several industries are increasingly investigating and adopting to help modernize their operations. Here are a few principles that UNS employs to achieve its goals:

To help with scaling, UNS recommends avoiding point-to-point connections between systems. For example, in a typical industrial scenario, a layered approach is a common design where data moves upwards from sensors -> PLCs -> SCADA -> MES, with commands usually moving in the opposite direction. Instead of point-to-point connections, data is moved to UNS from each layer directly. This hub/spoke topology is more scalable, especially when implemented using a publisher/subscriber component.

Another key aspect of UNS is federation, which allows the framework to span across multiple physical constructs, often split at hierarchical levels, such as edge UNS and enterprise UNS. This enables scaling out of UNS and restricting data access at the physical level, enabling the necessary isolation of resources and data often required in industrial environments.

Finally, data normalization is crucial in a UNS to provide a consistent view of operations. When connecting to various devices and controllers at the edge, it is highly likely that field names and units of measurement differ, even for the same type of sensors. A processing layer is required to map vendor/factory-specific fields to canonical fields and standardize units of measurement, enabling accurate data aggregation and inference. Further, the reported data must be placed within a consistent and intuitive hierarchy to facilitate easy discovery and consumption by other participating systems.

How Microsoft can help

Microsoft's adaptive cloud approach provides key building blocks that can be used in moving towards enterprise-ready UNS. Several contemporary UNS implementations leverage a Message Queuing Telemetry Transport (MQTT) broker running at the edge, such as the one provided in Azure IoT Operations, as the collection point for all site-level operational data. The broker aggregates data from disparate sources, like industrial control systems and equipment on the shop floor, Manufacturing Execution Systems (MES), etc. Data is then typically organized hierarchically using custom MQTT topics in a standard way and made available in real-time to consumers. In addition to a modern, scalable, and MQTT specification-complaint broker, Azure IoT Operations offers key capabilities like asset discovery, connectivity, and data processing to ingest and normalize operational data, therefore providing a solid foundation to build UNS.

With any UNS, the true value lies not only in data acquisition and normalization at the site level, but also in integrating this data into a holistic data platform strategy that can be used to gain insights across the entire business. Microsoft Fabric can provide a centralized platform in the cloud for insights and analytics from disparate OT and IT data sources. This goes beyond what UNS provides at individual factories. Using Fabric, companies can access tangible benefits like reduced energy consumption, increased production yield, enhanced product quality, and progressing sustainability goals.

In summary, UNS is a key strategy that organizations can implement to unlock value from industrial data. Azure’s adaptive cloud approach, including Azure IoT Operations and Microsoft Fabric, complements the use of UNS by enabling a comprehensive data platform strategy that can bring OT and IT data together through common models. It addresses the challenge of schema integration that UNS alone does not solve and provides an enterprise-ready product experience for the entire data insights journey. This positions organizations to capitalize on the full value of their data assets across various business domains, driving true digital transformation.

Learn more

Try out Azure IoT Operations Preview:

https://azure.microsoft.com/products/iot-operations