Human-Centered Artificial Intelligence for Quality 4.0: Adaptive and Generative Approaches to In-Process Manufacturing Quality

This PhD project focuses on human-centered artificial intelligence for Quality 4.0 in production systems. In the context of Industry 4.0 and Industry 5.0, the objective is to improve in-process manufacturing quality monitoring by combining sensor data, operator feedback, and adaptive and generative AI mechanisms. Classical quality-control methods, such as Statistical Process Control or Six Sigma, often detect problems after defects have appeared, while many AI-based solutions remain difficult to adapt to variable industrial environments and insufficiently interpretable for operators.
The thesis aims to design a human-centered Quality 4.0 framework in which operators are not only data providers, but active contributors to validation, explanation, and continuous system improvement. Adaptive learning loops, based on reliability scoring, transfer learning and possibly incremental learning, will be used to adjust models to changes in the production context. Generative AI will be investigated as a mediation interface between algorithms and human users, with the aim of providing contextualized explanations, supporting conversational validation of anomalies, and increasing trust in AI-assisted decisions.
The methodology will first include a systematic literature review on human-centered AI, Industry 5.0, explainable AI and generative AI in manufacturing quality. It will then involve the development of prototypes combining real-time data streams, adaptive defect detection algorithms and a generative explanation interface. These prototypes will be evaluated through simulations, controlled experiments with operators and, where possible, industrial case studies. Evaluation criteria will include detection performance, transparency, operator trust, feedback quality, workload, and accessibility for SMEs.
The expected contributions are conceptual, technical, empirical and practical. The thesis will propose a framework that integrates generative human-AI interaction as a pillar of Quality 4.0, develop adaptive and conversational quality-monitoring prototypes, provide empirical evidence on how generative explanations influence trust and feedback quality, and formulate recommendations for the sustainable adoption of lightweight human-centered AI solutions in manufacturing SMEs. 

Keywords: Human-centered artificial intelligence,Quality 4.0,Transfer learning,Generative AI,Quality monitoring,Production systems