How AI Startups are Changing Manufacturing

On a racetrack, cars aim to go fast - very fast - and, if possible, faster than everyone else. The effort to achieve that starts with the car’s design: inside wind tunnels and simulation labs. In Formula 1, until recently, testing aerodynamics meant running simulations that took hours or even days. Now, a British startup called BeyondMath ran comparable analyses in under 100 seconds. Their “digital wind tunnel” uses artificial intelligence (AI) to analyze thousands of variables at record speed, shortening development cycles and enabling more ambitious designs.

Yet, as impressive as this is, it’s not just about racing. It’s about how AI is beginning to transform the way entire industries design, plan, and produce goods. For years, AI in manufacturing was talked about more than implemented. That’s changing. As data is involved in every step of the value chain, AI can use it to inform better decisions – decisions that make manufacturing more adaptive, efficient, and sustainable.

The founders of BeyondMath from Cambridge, Britain, understand that as well. That’s why, in spring 2025, they attended the Siemens AI with Purpose Summit in Munich – to network, share ideas, cook up new projects and, alongside other startups, present their work to industry stakeholders.

Events like this summit are an excellent opportunity for startups, as they cannot succeed alone in getting their product out – especially in industrial manufacturing, where tasks are too complex, too interconnected. Yet, with the help of others, including startups and established players (such as companies like Accenture, SAP, and Infineon, that also attended the summit), success becomes more attainable. It’s even easier if it happens within an interoperable ecosystem of solutions. Which, just by chance, is what Siemens Xcelerator Marketplace is designed to provide.

Then there’s Blockbrain, which supports people on the shop floor in daily decision-making. It does so by capturing all the expertise that is locked in the minds of experienced workers - and lost when they leave.

While it resembles an LLM, it´s trained on in-house, non-public procedures, logs, and multimodal data. Imagine a technician facing an unfamiliar error code: instead of flipping through manuals or waiting for an expert, the program suggests step-by-step guidance based on prior fixes. This way, the company builds a living, growing knowledge base that supports consistent decision-making.

Individually, these startups tackle different challenges. Together, they point toward a more connected, data-driven approach to running factories - where design, planning, and operations are parts of the same digital ecosystem. That vision aligns with Siemens Xcelerator and the shift toward interoperable, modular industry solutions.

As part of this journey, the Siemens for Startups program helps founders accelerate their innovation by providing easy access to industry‑grade software, by offering co‑development opportunities where Siemens acts as an early customer to help turn ideas into reality, and by connecting their solutions to our global marketplace to help them scale.

As sellers on the Siemens Xcelerator Marketplace, they can place their tools alongside complementary offerings, making it easier for customers to test, combine, and deploy them. They access Siemens’ global industrial audience, common interoperability standards, and co-innovation opportunities. All of this will accelerate the adoption of industrial AI in the manufacturing sector. Ultimately, we end up with factories that move quickly – not as fast as Formula 1 cars, but fast enough to manufacture efficiently without unnecessary waste.