Hannover Messe 2026 again showed why it remains one of the world’s most important industrial stages. The fair was attended by the German Chancellor and several federal ministers, including the Federal Ministry for Economic Affairs, a key stakeholder for initiatives such as RoX, the digital ecosystem for AI-based robotics, and 8ra/ IPCEI-CIS, Europe’s major cloud-edge initiative. For us and our new AI-Data team, Hannover Messe was therefore more than a trade fair. It was a live checkpoint on how quickly AI, data ecosystems, and industrial policy are beginning to converge.
Figure 1: Briefing for Telekom C-level and government leaders of Industrie 4.0 initiative
Telekom C-level and government leadership at RoX
A highlight for us was the visit by Deutsche Telekom and T-Systems leadership to the RoX booth (see Figure 1), where Chris Langdon briefed them in his role as Telekom’s principal investigator in RoX and within RoX as lead for the data-ecosystem layer where he is running the consortium’s two dataspace environments (one for development & testing, the other for production; see Figure 3 for details) both subcontracted from TSI and provided as Dataspace-as-a-Service (DaaS, TSI’s Build & Operate product). RoX is a digital ecosystem for AI-based robotics that combines applications, data products and digital twins, and a dataspace network into an industrial environment for better, more flexible robotics. It shows physical AI in action today with three live robotic demonstrators on site. Just as important, these apps sit on top of a data ecosystem enabled by modern Web3-oriented decentralized dataspace technology, effectively an “Internet for data” that allows data to be shared and used with governance, trust, and sovereignty (Guggenberger et al. 2025). RoX also has strategic weight because it brings together an unusually strong consortium across both robotics and AI-data. On the robotics side, this includes leaders such as ABB, Dürr, Rheinmetall/ Yardstick together with pioneers like Sotec and Roboception. On the AI-data side, it includes organizations such as DFKI, DLR, Google’s Intrinsic, and T-Systems, giving the effort both industrial depth and access to frontier work in AI-data enabled robotics.
Figure 2: Executive roundtable – From Physical AI to industrialization with agents and dataspaces
Executive roundtable: From Physical AI in action with RoX to industrialization with agents and dataspace
A key activity was our executive roundtable on Physical AI with RoX (see Figure 2). The underlying logic was straightforward: AI-data readiness has now moved from the IT department into the boardroom (Davenport & Tiwari 2024). Senior leaders increasingly understand that AI success depends not only on models, but on governed access to the right data, including sensitive and proprietary data across company boundaries and global supply chains. RoX created a rare setting for such a discussion. Because the project is already a real industrial effort with three live robot cells running on the showroom floor, the conversation did not need to stay at the level of vision or speculation. Instead, the roundtable could focus on first lessons learned, implementation realities, and emerging best practice. Chris Langdon designed and moderated the roundtable, using the Drucker School’s heritage and Peter Drucker’s management authority as a neutral top-management platform. Combined with his own AI-data publications, this created a setting for exchange and debate that was both rigorous and relevant. The format builds on earlier C-suite roundtables:
- First at CES 2026 hosted by T-Mobile
- Then at Automotive Industry Action Group (AIAG) Detroit HQ in April. Both focused on AI-data readiness through data ecosystems, including Catena-X, the first open and productive data ecosystem in automotive
- Hannover Messe was the next logical step. The format expanded beyond automotive into real-world Physical AI examples, with the first live AI-data-enabled robotics use cases. This allowed the roundtable to be paired directly with live demonstrations on the showroom floor, because seeing is believing
Figure 3: 3-layer data ecosystem stack and RoX “middleware” roadmap for physical AI
Same tech stack, two directions: (1) global data ecosystems and the (2) enterprise dataspace
A central message from Hannover Messe was that the same underlying dataspace stack (see Figure 3 for 3-layer software stack) now supports two strategic directions at once: (1) global, cross-company data ecosystems and (2) enterprise dataspaces inside companies.
(1) Global data ecosystems
Catena-X in automotive has emerged as the most prominet data ecosystem, where the dataspace network layer is already live and in production. On top of that foundation sit applications such as traceability and product carbon footprint, as well as the related data products and digital twins or aspect models. This matters because it shows that the concept has moved beyond prototyping and pilots into an operating environment. As of 2026 Cofinity-X is the sole operating entitiy of a Catena-X data ecosystems with hundreds of registered users and already tens of thousands of peer-to-peer data transactions per day (see Figure 3 left side for Catena-X data ecosystem).
Figure 4: Collage of our interoperability activities involving EU-Japan projects, sessions at OPC Foundation and IDSA, and visit by China Telecom delegation
NTT DOCOMO Business–Fujitsu–DENSO–T-Systems: Federated ID & trust anchor solution. At Hannover Messe, we demonstrated how this logic can be extended across regions through EU-Japan dataspace interoperability (see Figure 4 for global projects). Together with NTT DOCOMO Business, Fujitsu, DENSO, and T-Systems the demonstration showed that trusted cross-border data exchange requires more than connectivity. It also requires a cross-juridiction mechanism to verify that participating companies are trustworthy and that the data being exchanged actually comes from them. On the Japan side, this involved connecting into gBizID’s trial environment, Japan’s corporate authentication infrastructure. NTT DOCOMO Business verified corporate identity in the testbed, Fujitsu issued the relevant dataspace credentials, DENSO provided the Battery Passport application, and T-Systems provided the certified dataspace connectivity and operating environment. The result was a practical demonstration of how data ecosystems can begin to follow global supply chains rather than stopping at national or regional borders.
Visit by China Telecom: Global interoperability drew strong outside interest, including a visit from China Telecom. That matters strategically because once data ecosystems begin to connect across regions, the discussion moves from one industry initiative to the broader question of how a global “data dial tone” can emerge under conditions of sovereignty and trust.
Figure 5: HM2026 was top level event with key partners focused on our AI-data solutions
(2) Enterprise dataspace
The second direction is as important as global scope: applying the same technology stack inside the enterprise. The logic goes back to Peter Drucker, who identified the core constraint decades ago: “economic-chain costing requires information sharing across companies, and even within the same company, people tend to resist information sharing” (Drucker 1995, p. 58). That constraint is now becoming critical for AI. If GenAI and agentic systems require relevant, auditable, and governed data, companies must first address the fragmentation of their own internal data landscapes and enable data sharing across silos (Davenport & Tiwari 2024). This is where the ‘enterprise dataspace’ comes in. Rather than replacing existing systems, it creates a governed layer across them. It allows companies to connect internal silos, expose governed data products, and support AI use cases without surrendering control. This is not theory. As shown in the attached IAV case, the same open-source Eclipse Tractus-X stack used in Catena-X can be turned “inside out” as an internal enterprise dataspace. In practice, this helps close the AI-data-readiness gap while supporting the broader concept of the ‘extended enterprise’: collaboration that remains governed inside the company, yet is structurally ready to extend across partners, suppliers, and ecosystems when needed.
Key takeaways
(1) AI-data readiness is now a C-level issue and economic-policy affair. That was visible not only in executive attention and booth traffic, but also in the broader ministerial and policy context of the fair. Once AI depends on supply chains, raw materials, industrial resilience, and trusted cross-border data exchange, it is no longer a niche IT topic.
(2) Physical AI needs more than models, it needs the right information content. Large language models and other AI systems only deliver reliable results when they are fed with relevant, governed, and often proprietary data. Or more bluntly: no cats in the training data, no cats out of it. Dataspace technology is emerging just in time as the practical answer, because it provides the governance layer needed to unlock the right data without losing control of it.
(3) The same Web3-oriented, decentralized dataspace technology now works in two directions. Externally, it powers global data ecosystems such as Catena-X. Internally, it follows the historical pattern of telephony and the Internet: technologies that first emerged as external networks and were later adopted inside the enterprise. Applied this way, the technology enables an enterprise dataspace as the foundation for the extended enterprise: a CEO-level response to AI-data readiness, regional fragmentation, and supply-chain resilience. It allows companies to stay in control internally while becoming structurally ready to collaborate across partners, suppliers, and ecosystems as needed.
Deep dive: Insights, lessons learned, business impact
- Dataspace-CEO-2-pager: What is … data ecosystem … dataspace … 3-layer stack?
- Langdon’s Drucker-based roundtables: CES 2026 hosted by T-Mobile, AIAG HQ April 2026
- Case study: The ‘extended enterprise’ strategy powered by ‘enterprise dataspace’
References
Drucker, P. 1995. The information executives truly need. Harvard Business Review (January-February; winner of the HBR McKinsey Award for the year’s best HBR article): 54-62
Davenport, T. H., and P. Tiwari. 2024. Is your company’s data ready for generative AI? Harvard Business Review (April), link
Guggenberger, T. M., C. Schlueter Langdon, and B. Otto. 2025. Data Spaces as Meta-Organisations. European Journal of Information Systems, January: 822-842, link
