PhD in Process-Aware AI Agents for Autonomous Manufacturing at Process Analytics Cluster

PhD position in Process-Aware AI Agents for Autonomous Manufacturing at the Process Analytics Cluster, Department of Mathematics and Computer Science at TU/e. Part of the EuroFMX project (Horizon Europe initiative with 69 partners from 20 countries) developing the world's first industrial-native autonomation foundation model for manufacturing using Physics-Informed Graph Neural Networks and transformer architectures over knowledge graphs.

Research directions include:

• Ontology-based construction of event knowledge graphs from heterogeneous industrial data sources
• Architecture for agentic AI operating over event knowledge graphs
• Active inference agents with graph-based, physics-informed generative models
• Industrial validation of agentic AI methods with partners like Philips and Continental

The position involves collaboration with BIASlab (Electrical Engineering) on active inference and Celonis on graph-based generative AI tools. Work will be validated in real-world manufacturing use cases spanning automotive, aerospace, and electronics. Four-year full-time employment with minimum 10% and maximum 15% per year teaching tasks. International consortium collaboration expected with travel and participation in project meetings.