Last month, I wrote about the ethical use of AI in our daily lives. I touched on the use and benefit of AI within the industrial industry; but South Africa’s manufacturing industry, like many others globally, faces challenges in digital transformation. Legacy systems, siloed data and inconsistent communication protocols hinder our ability to extract meaningful insights from operational data. Engineers and analysts often struggle to access the real-time, contextualised information that is essential for predictive maintenance, quality assurance and autonomous decision making.
Traditional approaches, such as building custom APIs or creating OT data lakes, often introduce complexity, increase costs and fail to deliver the agility needed, even for modern AI systems. This month, I want to explore two concepts, backed by different architectures and technologies, that can assist our manufacturing facilities to exponentially increase the pace of their digital transformation. These are Unified Namespace (UNS) and Model Context Protocol (MCP).
UNS is a revolutionary concept that addresses challenges head-on. Rooted in MQTT and Sparkplug B protocols, UNS acts as a real-time, centralised data fabric for industrial operations. The concept has been around for some time, but only in recent years has it found a popular name.
UNS is a publish-subscribe architecture that organises industrial data such as sensor readings, machine states and production metrics into a hierarchical, standardised structure. It democratises data access across sites, departments, systems and applications, enabling seamless integration between OT and IT layers. By implementing UNS, manufacturers create a single source for all operational data, an essential foundation for AI readiness. However, UNS only contains the current state, measurement or value of any defined data point, with no history. Historical data, if important, still needs to be accessed elsewhere, either at the source or from the subscriber.
While UNS provides the backbone, MCP is the enabler, serving as the standardised interface, allowing AI models to interact effectively. MCP servers curate and contextualise data from UNS, delivering it in a format AI agents can understand and act upon. MCP offers standardised access, context persistence, interoperability and scalability. UNS and MCP are a powerful combination. UNS organises and contextualises data, while MCP enables intelligent systems to consume and act on it in real time.
A vision for South Africa’s manufacturing sector
South Africa is home to a diverse and resilient manufacturing base, from automotive and electronics to food processing and chemicals; but to remain competitive in the global arena, we must embrace digital transformation, not as a buzzword, but as a strategic imperative. As described in the President’s vision for 4IR, South Africa can use the innovative nature of our population to accelerate adoption of 4IR technologies such as AI to improve the competitiveness of our industrial sector. Especially with the imposed tariffs from the USA, our country needs to become more competitive so we can compete on the global market.
The more competitive our factories and manufacturing plants, the better we can compete and win business from the other, lower cost regions in the world.
The advent of commercially available AI solutions represents a once-in-a-generation opportunity to redefine how we build, operate and optimise industrial systems. It is the key to unlocking the full potential of digitisation in manufacturing, and to positioning South Africa as a leader in the global industrial AI revolution. This transformation is not just technical, it‘s cultural. It requires collaboration between academia, industry and government. It demands investment in skills development, infrastructure and change management; and it calls for leadership, something the SAIMC is uniquely positioned to provide.
Call to action
A good start for anyone is the SAIMC User Conference in October 2025 where various speakers, international and local, will share their knowledge and success stories in digitising factories. There will also be interactive panel discussions with industry leaders to discuss how South Africa can benefit from technology in manufacturing facilities. Attending this conference will expose attendees to the ‘art of the possible’, where knowledge can be shared and options discussed.
I urge all members, readers and interested parties such as academics, system integrators, OEMs, business owners and plant managers, to begin exploring digital tools within their organisations. Start small: pilot a digital automation implementation to monitor equipment efficiency, or a tool to track inventory movement and consumption. Experiment with MCP servers to connect AI models to your production and inventory data. Share your learnings, successes and challenges with the community.
Let us also advocate for national initiatives that support smart manufacturing through policy, funding and public-private partnerships. South Africa has the talent, the infrastructure and the ambition. What we need now is alignment and action.
| Tel: | +27 11 312 2445 |
| Email: | [email protected] |
| www: | www.saimc.co.za |
| Articles: | More information and articles about SAIMC |
© Technews Publishing (Pty) Ltd | All Rights Reserved
printer friendly version