For years, manufacturers have experimented with Industry 4.0 — running pilots in predictive maintenance, production monitoring, and AI-driven optimisation.
Yet many of these initiatives struggled to move beyond proof-of-concept. The underlying problem wasn’t a lack of ideas, but the difficulty of accessing and structuring machine data across complex factory environments.
United Manufacturing Hub Systems (UMH) is building an open-source data platform designed to solve exactly that challenge. The company now counts manufacturing, food and beverage, and top-20 automotive suppliers among its customers, supporting deployments across more than 150 sites globally.
Oh, and the company raised €5 million in January.
I spoke to Nikklas Hebborn, CCO of UMH, to learn more.
From venture investor to operator
Before joining UMH operationally, Hebborn was a partner at Freigeist Capital, the German deep-tech venture firm founded by Frank Thelen. His background spans venture capital, consulting, and advisory roles with organisations including Bayer, Capgemini Invent, Roland Berger, and Kearney.
Hebborn later transitioned from investor to operator, moving from backing United Manufacturing Hub as an early investor to joining the company’s leadership team.
Founders who experienced the problem firsthand
The company was founded by CEO Alexander Krüger and CTO Jeremy Theocharis. After graduating from RWTH Aachen University, the pair worked on digitalisation projects for large consultancies such as McKinsey, travelling globally — from Tokyo and Singapore to Atlanta — deploying industrial use cases directly on factory shop floors.
Over several years of doing this work, they noticed a recurring challenge. Building the actual use case — whether a dashboard for energy monitoring, productivity tracking, or even deploying AI — represented only about 10 per cent of the work. The remaining 90 per cent involved collecting and preparing the data: gathering it in real time, ensuring it had the correct format and context, and maintaining sufficient quality.
“That’s where most projects struggled,” explained Nikklas Hebborn.
The real bottleneck in industrial digitalisation
The founders realised that the real bottleneck wasn’t the applications themselves but the infrastructure required to reliably access industrial data. They therefore focused on building the underlying layer that connects operational technology (OT) — machines and sensors on the factory floor — with IT systems such as ERP platforms.
The result was a platform built around what the company calls a Unified Namespace: a structured data environment that allows companies to move seamlessly from a site-level overview down to individual machines or even specific sensor readings.
According to Hebborn, the founders’ deep industry experience was critical to shaping this approach.
“Many startups identify a problem and then bring in domain expertise later,” he said.
“Here it was the opposite — the founders lived with the problem for nearly a decade.
They had personally experienced the pressure of delivering digitalisation projects under tight timelines while working on shop floors around the world. That deep understanding of the problem space was what convinced me to join.”
UMH’s platform helps manufacturers collect and structure data from machines, sensors, and factory software systems. Modern factories run a mix of legacy equipment, industrial controllers, and enterprise software, all producing data in different formats. The platform gathers machine data via common industrial protocols and transforms it into a unified, real-time stream that feeds dashboards, analytics tools, manufacturing execution systems (MES), or AI models.
A key concept behind the platform is the “Unified Namespace,” which acts as a single source of truth for factory data. Instead of each application pulling information separately from machines or databases, data is published once into a shared structure that authorised systems can access. This simplifies integration, improves transparency across production processes, and accelerates Industry 4.0 use cases such as predictive maintenance, energy optimisation, and production monitoring.
Under the hood, the platform is designed as a modular infrastructure layer with tools to manage deployments across factories. UMH’s solution has two main components.
The first is the infrastructure layer, configured through code via a large configuration file. On top sits a management console that acts as a control centre for deploying instances, connecting machines, building data bridges between systems, and defining data models. Hebborn explained:
“One automotive supplier we work with has CNC machines worldwide. Each machine produces millions of data points, but only a handful are actually relevant. The raw outputs are often cryptic values like ‘DB3456’, which might represent temperature or pressure.”
Within the console, companies create data models that translate these signals into contextualised, understandable information.
The platform supports two interaction modes. Developers can configure deployments through YAML files — something Hebborn says AI tools can generate quickly when connecting hundreds of machines. For non-technical users, UMH also offers a visual drag-and-drop interface, which becomes important when deployments scale across dozens of sites.
The value of open source and interoperability
A defining decision of UMH was to build the platform as open source, an unusual move in an industry dominated by large incumbents.
“There are many big players here, from Siemens and Rockwell to the hyperscalers on the IT side trying to move industrial data into the cloud,” said Hebborn.
“But customers increasingly want to own their software. If you want to stay competitive and become something like a Tesla in manufacturing, you need to be software-driven and vertically integrated. Open source allows companies to maintain that ownership.”
The second reason, he explained, is interoperability. “
A typical factory floor might use Siemens machinery, Rockwell automation, and Microsoft cloud infrastructure. Companies need an independent layer in the middle that connects all of these ecosystems.
To achieve that, the founders built its platform around an open-source model and cultivated a broad community around it. Today, more than 1,000 system integrators, consultants, and end users are using the community edition.
“That effectively gives us hundreds of people constantly testing the product, identifying issues, and contributing feedback,” Hebborn said.
“Many competitors develop a product internally and only receive feedback once it’s deployed at a customer. Our development cycle benefits from a much larger real-world testing base.”
That community-driven approach has also helped drive organic adoption. “Interestingly, many customers actually discover us themselves,” he added.
“Often, they first attempt to build their own solution internally. Eventually they realise how difficult it is to scale and then start looking for an infrastructure layer like ours.”
Why the real challenge wasn’t the application
Industry 4.0 and Industrial IoT have been discussed for years, but many companies have run pilots that have struggled to scale, and no dominant platform has emerged. According to Hebborn, what has changed over the past five to ten years is the availability and accessibility of machine data.
“Ten years ago, getting data from industrial machines was difficult. Vendors like Siemens built closed ecosystems.
Today, there are far more standardised protocols and APIs available to extract that data. Another shift is infrastructure.
Production sites now have much better IT connectivity, including direct internet access on the shop floor. And finally, companies have already experimented with digitalisation.
Many consultancies sold digital transformation projects that led to numerous pilots and proof-of-concept use cases. Many of those didn’t scale, but the demand didn’t disappear.”
Today, Hebborn sees companies with a“chessboard” of use cases they want to implement. “They know the potential is there—they just lacked the underlying infrastructure layer to do it properly.”
A key decision for the company was focus. Its CTO has a strong opinion about this: we want to be the best data layer, not the best tool for everything, explained Hebborn.
“For example, we don’t try to build our own visualisation tools. Customers can use platforms like Grafana, Power BI, or Snowflake.
We simply ensure the data is structured and accessible so those tools can use it. If we tried to build every component ourselves— visualisation, historians, analytics — the product would become too complex and we would risk turning into a consultancy.”
Hebborn stresses that you can’t bluff in a factory environment.
“One example was HiPP, the infant nutrition company. When we first installed the system and showed them the dashboard, the management team asked, “Is this really our data on the screen?” They had never seen their production data presented in that way before."
He found this surprising.
"We weren’t building something as complex as an electric vehicle, we were simply connecting data that already existed on the shop floor. But clearly, this problem still hasn't been solved properly.”
Another important element for UMH is training the customer team. It follows a “train-the-trainer” model where it trains one production specialist who then trains colleagues across the site and other facilities. Hebborn shared:
“In one case we didn’t hear from the team for two or three months. But we could see them continuously developing new use cases internally — and eventually they expanded the deployment to additional sites.”
Hebborn is modest about the company's success, sharing that while the company may not have hundreds of customers, "every customer we do have has expanded their deployment — and they’ve typically done so within twelve months."
"In manufacturing, that is extremely fast. If we eventually reach three or four hundred companies and scale across their sites, that would already represent a very large business. We don’t need thousands of customers to build a major company. That’s the proof point we’re most proud of today.”
UMH's next phase is about scaling through team expansion. Geographically, the team is not aggressively pushing international expansion yet as the DACH region already has a huge concentration of global manufacturing companies.
“Many of them operate internationally, so once we deploy locally, the solution often spreads across their global sites,” shared Hebborn.