Inside the Industrial AI Summit 2025: What Europe’s Leaders Agreed Must Happen Next

On September 4, 2025, more than 60 senior executives, policymakers, and AI researchers gathered at Schloss Donaueschingen for the Industrial AI Summit 2025, co-hosted by SPREAD, Pelico, and EthonAI, fellow members of the EU AI Champions Initiative. The discussions aligned around one central theme: How can Europe accelerate industrial AI adoption before its competitive edge fades?

Across keynotes, case studies, and panel conversations, the message was consistent: Europe has the engineering depth, manufacturing heritage, and research excellence to lead, but its ability to scale AI across production and product development remains too slow.

Europe’s AI Reality Check: Why 97% of GenAI Pilots Fail

In the opening keynote, Prof. Dr. Torbjørn Netland, Head of Production & Operations Management at ETH Zürich, delivered one of the most referenced insights of the day:
97% of GenAI pilots still fail to deliver ROI.

His explanation was direct.
GenAI struggles in industrial environments not because the algorithms are insufficient, but because data foundations are weak, fragmented, and lack context.

Netland warned that without fixing the underlying data problem, organizations will stay stuck in pilot purgatory, unable to reach the performance frontier that global AI leaders are beginning to access.

Engineering the Products of Tomorrow: Volkswagen & SPREAD

Frank Ortmann, Delivery Manager ID.Family, Volkswagen, and Philipp Noll, Co-Founder & Managing Director, SPREAD AI addressed the rising complexity of software-defined vehicles (SDVs). 

Ortmann explained how years of disconnected engineering systems have made it increasingly difficult to manage dependencies across hardware, software, and electronics and showcased how data-driven engineering can ‘shift left ’ in R&D to prevent late-stage failures that drive cost and delay in SDVs. By accelerating root-cause diagnosis, compressing validation cycles, and cutting time lost in fragmented systems. The broader lesson: data integration is the decisive factor in making industrial AI scale.

Panel Discussion: Europe’s Industrial Edge in the Age of AI

The panel session moderated by Vendeline von Bredow (The Economist) brought together an impressive lineup of speakers, including Sarah Engel (TRUMPF), Jonathan Barry (Mila), Robin Dechant (General Catalyst), and Jeffrey Hojlo (IDC).

Their discussion focused on Europe’s structural position in the global AI landscape. The panel agreed on several critical points:

1. Europe’s industrial base is strong, but digitization gaps persist: World-class engineering and manufacturing capability is not matched by the speed of AI integration into daily operations.
2. Scaling remains the core challenge: Industrial AI succeeds only when connected across engineering, production, quality, and service, a transition many companies have not yet achieved.
3. Capital availability matters: Compared with the US and Asia, Europe still underinvests in industrial AI scale-ups, limiting the speed at which technology diffuses across the manufacturing industry.
4. Trust and transparency can become strategic advantages: Europe’s regulatory and ethical frameworks can help accelerate responsible adoption, but only if they enable, rather than delay, operational deployments.
5. Cultural readiness is as important as technical readiness: Leadership alignment and cross-department collaboration remain essential for any successful AI scaling initiative.

The conversation concluded with a clear insight:
Europe still holds a competitive advantage, but it will diminish quickly if AI is not scaled systematically and with urgency.

Scaling Industrial AI in Manufacturing: Lessons from Siemens & EthonAI

The session on scaling AI in production featured Julian Senoner, CEO & Co-Founder, EthonAI, Sébastien Bey, CIO, Siemens Smart Infrastructure, Alexander Dierolf, Head of Data Intelligence & Automation, Siemens Smart Infrastructure

Their discussion on “Scaling Industrial AI in Manufacturing” broke down exactly how large organizations can move beyond pilots and into global deployment.

Bey described Siemens’ early challenges: high-mix production environments, inconsistent quality data, and visual inspection workloads exceeding 20,000 inspections per day.
Common solutions failed, not because AI didn’t work, but because they weren’t scalable.

Senoner outlined a proven four-step scaling motion:

1. Start with a real operational pain point

2. Demonstrate impact bottom-up

3. Replicate factory-to-factory

4. Secure global sponsorship

To conclude the session, Dierolf emphasized that success depended not only on technology but on empowering operators, not replacing them.

What Europe Must Do Next

The Industrial AI Summit 2025 made one thing unmistakably clear:
Europe is not behind in capability, it is behind in speed.

The expertise is here. The ecosystem is here. The will is emerging.
But impact will only come when organizations shift from fragmented pilots to coordinated, data-driven transformation across the entire value chain.

''Europe has all the ingredients to lead the industrial AI revolution, world-class engineering, trusted ecosystems, and the courage to transform. The next step is connecting these strengths into scalable systems that deliver impact across industries.''

As the summit showed, the path forward is not theoretical. It’s already being built, in Volkswagen’s engineering operations, in Siemens’ factories, in the growing rise of AI strategy in factories, and across the bold and innovative European ecosystem represented by SPREAD, Pelico, EthonAI, and the EU AI Champions Initiative.

Europe’s next chapter in industrial leadership depends on what happens now:
fix the data foundations, scale horizontally, and commit to the ecosystem that will carry Europe into the next era of manufacturing excellence.