← All articles

· Rimon Soliman

OT Data Fabric and UNS: The Data Foundation Before AI

A Plant Engineering article dated November 17, 2026 argues that an OT data fabric, with a Unified Namespace at its core, is the foundation layer for making industrial data AI-ready and for harmonizing SCADA, MES and historians.

Industrial AIUnified NamespaceOT data fabricSCADAMES

On November 17, 2026, Plant Engineering published a technical piece in its Digital Transformation section, written by John Hall of Novaspect. The thesis is clear: before building artificial intelligence applications, manufacturers need an OT data fabric, and within it a Unified Namespace (UNS), as the foundation layer for plant data.

The definitions used in the article

According to the author, a UNS is a structured, real-time data layer that organizes and exposes the current state of production systems within a consistent hierarchy. An OT data fabric, by contrast, is a data integration and management platform that connects shop-floor operations with enterprise and cloud systems and delivers aggregated, contextualized data. The UNS is therefore treated as one component of the data fabric, not as a synonym for it.

The article describes the UNS as an enterprise-to-control-system integration framework based on ISA-95 or a similar model. Its value, the author suggests, lies in covering all types of OT data, from every available source and across multiple plants. Better visualization would follow, at both site and enterprise level.

Three benefits attributed to the data fabric

The author lists three main advantages:

  • Harmonization across OT software platforms and systems of record, such as SCADA, MES, historians and connected-worker applications. According to the article, the need grows as the OT software stack proliferates and these applications spread.
  • AI readiness and robustness: the data fabric is presented as a necessary element.
  • Diagnostic and security monitoring at enterprise level, across every data connection and all data movement.

The context is ongoing digitalization, which the author expects to continue with goals such as higher yield, lower costs and waste, higher OEE, lower energy consumption, and faster, more accurate decisions.

How to read these claims

A few caveats apply. The article appears to be written from a vendor's perspective, so its claims reflect the author's point of view. In the portion I was able to read, no figures or case studies are cited. In addition, the page was truncated after the first section: the points on AI readiness and security monitoring rely on the article's bulleted summaries, not on a full reading.

That said, the argument is consistent with what is often seen in industrial projects: models and algorithms depend on the quality, consistency and context of the data. If the same asset has different names, structures and units in SCADA, MES and the historian, every AI initiative risks starting with cleanup and mapping work repeated from scratch.

First the contextualized data layer, then the AI layer.

Why it matters for automation professionals

This topic sits at the level of OT data architecture, distinct from pilot projects that stall before production or before the decisions to be supported have been mapped. It concerns the point where SCADA, historians and MES converge: integrating them is often the real groundwork of an AI project.

Practical takeaways

For those evaluating AI initiatives in the plant, the article suggests a few starting questions:

  • Is there a shared asset model and hierarchy (for example, inspired by ISA-95) used by all systems?
  • Which OT sources are already accessible in a consistent way, and which remain isolated in individual plants or applications?
  • Who monitors the state of data connections and flows between shop floor, enterprise and cloud?
  • Is the data model replicable across multiple sites, or does it have to be readapted for each plant?

The operational conclusion is simple: build the UNS and data fabric before the AI layer, while critically assessing vendor promises and verifying them against concrete cases in your own plant.