TEAMING. AI Final Conference | Follow-up

The final conference of TEAMING. AI, titled "Managing Human-AI Collaborations within Industry 5.0 Scenarios via Knowledge Graphs - A Summary of TEAMING. AI Project’s Outcomes and Lessons Learned," took place online on May 23rd. The event focused on knowledge graph-based technologies and approaches that enable the management of human intervention in AI-assisted manufacturing processes in Industry 5.0, especially under potentially changing conditions, to maintain or improve overall system performance.

While knowledge graph-based systems are typically designed with a static structure, the European project TEAMING. AI addresses the dynamic challenge of inline human-AI collaboration in industrial settings. In this context, we discussed approaches and lessons learned, addressing general challenges such as modeling domain expertise with a particular focus on vertical knowledge integration. We also tackled challenges linked to an industrial knowledge graph, such as its dynamic population and the late shaping of knowledge graph embeddings. These embeddings serve as the foundation of relational machine learning models, which have emerged as an effective tool for exploiting graph-structured data to infer new insights. Additionally, we considered legal and ethical aspects from the perspective of European AI law.

More specifically, the sessions of the conference focused on:

• Requirements, human factors, and legal aspects

• Knowledge graph-based approaches to human-AI collaboration

• Application integration and proof of concept

• Exploitation potentials and lessons learned

Check out the agenda here. 

In case you missed it, you can now watch the discussion on YouTube here.