Deploying Real-World UNS Architectures with MQTT and Node-RED

As industrial companies race to modernize their data infrastructure, Unified Namespace (UNS) has emerged as a practical solution to bridge OT and IT systems at scale. In Episode 4 of CONNACK, Alejandro Simó Vesperinas, IT/OT Architect at Mayker, took the stage in Munich to walk us through a practical, field-tested approach to deploying Unified Namespace (UNS) architectures. His session spotlighted the power of MQTT and Node-RED, and how Mayker has developed custom tooling to streamline data modeling and OT/IT integration, ultimately delivering real value in industrial environments.

Alejandro’s talk covered the use of a scalable Node-RED deployment, lessons learned from real customer projects, and best practices for building maintainable data pipelines.

Watch the full talk below, and read on as we unpack the key takeaways from the presentation.

OT/IT Integration Challenges in 2025 and How to Solve Them

Before diving into architecture and tooling, Alejandro addressed some of the common pain points that manufacturers and industrial teams face when integrating OT and IT systems:

• System Silos: Legacy architectures often create isolated systems that don’t easily communicate, leading to duplicated data and effort.

• Manual Data Handling: Extracting and formatting industrial data is still a manual, error-prone process in many factories.

• Inconsistent Data Quality: Without standardized models or schemas, data lacks structure, making downstream analytics unreliable.

• Security and Scalability: IT/OT convergence introduces security risks and complexity, especially when scaling from pilot to production.

Traditional ISA-95 stack models enforce communication only between adjacent levels. But with IoT technologies, systems now need to talk across layers, creating the need for a new approach: the Unified Namespace.

The UNS provides a single source of truth for industrial data. Instead of point-to-point integrations, UNS offers a structured, hierarchical data architecture.

UNS and Data Modeling: Bringing Structure to Chaos

To make UNS truly valuable, the data flowing through it must be clean, contextualized, and meaningful, because ‘garbage in, garbage out’ is a reality in manufacturing.  

“

With the challenges of OT/IT convergence on the table, Alejandro showed how Mayker tackles them by using Node-RED at the edge to turn cryptic tags into structured, contextualized UNS data streams. Specifically, Mayker's approach emphasizes data modeling at the edge using Node-RED. This method ensures that:

• Data is contextualized close to its source.

• Network traffic is reduced by sending only structured, meaningful payloads.

• All downstream consumers, such as dashboards, analytics platforms, ERP systems, receive usable, high-quality data.

Why Node-RED?

For Mayker, Node-RED offered the ideal mix of rapid prototyping, ease of use, and the ability to contextualize data right where it’s generated. Here are few more points:

• Familiar & Flexible: Many engineers start using Node-RED at home and later bring it into work environments.

• Open-source & Rapid Prototyping: Ideal for agile development.

• Extendable with FlowFuse: FlowFuse makes Node-RED enterprise-ready by enabling version control, CI/CD pipelines, and centralized deployment management.

Mayker has even developed custom Node-RED nodes, such as:

• An OPC UA browser to easily select tags.

• A modular mapper to associate raw data with contextual models (e.g., conveyor, filler).

• Support for converting data into standardized JSON, which can then be published via MQTT.

What Enterprises Can Learn from Mayker’s UNS Approach

Mayker’s real-world deployments offer valuable lessons for enterprises looking to scale Unified Namespace architectures with confidence and clarity.

1. Model Before You Move

Raw industrial data is often cryptic. To be useful, it must be enriched with metadata and structure. Mayker's “instance-model-attribute” approach helps ensure clarity and consistency across systems.

2. Contextualize at the Edge

Embedding intelligence at the edge (e.g., via Node-RED) ensures context travels with the data. This reduces latency, saves bandwidth, and simplifies consumption for downstream systems.

3. Use Open Standards (But Be Practical)

Mayker recommends using ISA-95 as a reference for topic structures, and exploring CloudEvents to standardize event payloads. But flexibility is crucial, adapt schemas to fit your business case.

4. Manage at Scale 

FlowFuse enables scalable Node-RED deployments by allowing:

• Centralized instance management

• CI/CD for flows

• Governance over edge logic

This is key when managing hundreds of remote Node-RED instances across industrial sites.

5. Build for Business Value

Mayker's real-world projects started with clear business goals, from energy monitoring to ERP integration. Focus first on proof-of-value use cases that deliver measurable impact.

6. Create a Strong Data Culture

UNS success depends on more than tools. It requires:

• Cross-functional IT/OT collaboration

• Documentation and governance

• Upskilling teams

• Iterative rollouts and refinements

Real-World Deployments: From Theory to Impact

Alejandro shared three customer examples:

• Food Manufacturer: Retroffited sensors and integrated historian data into MES via UNS. Dashboards created in Grafana.

• Industrial Client: Solved governance challenges by deploying FlowFuse at the edge and integrating with ERP via MuleSoft.

• Energy Use Case: Used Node-RED for extracting historian data and improving energy KPIs with a fully modeled UNS structure.

Each project delivered tangible outcomes by focusing on data usability, not just data availability.

Final Thoughts: Building the Factory of the Future

UNS isn't just a buzzword. It’s becoming a foundational element in modern industrial data architectures. But implementing it well requires the right tools, the right approach, and the right mindset.

Alejandro’s advice to solution architects and digital leaders is simple but profound:

“Start small. Focus on impact. Build models. Govern your namespace. And always think about the consumer of your data.”

At HiveMQ, we’re proud to be the reliable backbone for many such architectures, ensuring that contextualized, high-quality data flows securely and efficiently across the enterprise.

Whether you’re integrating edge devices, building a digital twin, or scaling enterprise data pipelines, HiveMQ’s IoT Data Streaming platform helps make UNS architectures real and reliable.