Fraunhofer Institute for Software and Systems EngineeringIndustrial AI in the Ecosystem: Creating Value Together
Industrial companies are increasingly investing in AI solutions to leverage their data effectively and gain a competitive edge. However, AI’s full potential is only realized when models can learn from distributed data across industrial ecosystems—use cases such as predictive quality assurance or optimized supply chains only become truly feasible then.
Various approaches make this possible: data rooms provide the foundation for secure, sovereign data exchange between partners; federated learning enables the joint training of models without data transfer; and targeted unlearning ensures that a partner’s data can be completely removed from a shared model upon request. Control over one’s own data is always maintained.
Fraunhofer ISST and Fujitsu will demonstrate how these approaches work together in practice at Hannover Messe 2026. Visit us at the exhibit or join our masterclass—we look forward to the exchange.
Exhibit: How Can AI Forget?
“Decentralized Federated Unlearning in Zero-Trust Environments”: Fujitsu Research and Fraunhofer present a solution for decentralized federated unlearning without quality loss in zero-trust AI models.
Having a proprietary AI model isn’t practical for every company. Instead, federated learning enables a decentralized, jointly trained model with distributed rights. But what happens when a participant leaves the ecosystem and their data needs to be “unlearned”?
Fujitsu Research and Fraunhofer ISST demonstrate how secure unlearning can be achieved in zero-trust environments without significant losses in accuracy.
When: April 20 – 24, 2026
Where: Hannover Messe, Hall 11, Booth D33 / Fraunhofer Joint Booth
Contact: Florian Zimmer and Sebastian Becker
Masterclass: “Data Space Technologies as Enablers for Industrial AI” (Session in English)
Building on the latest research findings, this session explores technologies for utilizing data spaces in the training and inference of industrial AI models. It will also discuss the opportunities and challenges involved in developing large industrial models from both technical and economic perspectives.
When: Tuesday, April 21, 2026, 12:15 – 1:15 p.m.
Where: Solution Lab, Hall 26, Booth E43Masterclass Room 2
Speaker: Boris Otto (Fraunhofer ISST), Taka Matsutsuka (Fujitsu)
Registration for the Masterclass (only possible with a valid ticket)
Contact
Florian Zimmer Research Associate Industrial Manufacturing Fraunhofer-Institute for Software- and Systemengineering ISST |
Sebastian Becker Research Associate Mobility & Smart Cities Fraunhofer-Institut for Software- and Systemengineering ISST |