Integration of Artificial Intelligence and Total Quality Management in Manufacturing Projects for Industry 5.0

Industry 5.0 shifts the focus of manufacturing from mere productivity to sustainability, resilience, and humancentric value. This evolution demands that quality management systems (QMS) are not only effective but also explainable and auditable. This paper is a conceptual framework study that synthesizes TQM, Quality 4.0, trustworthy AI, and digital twin principles to propose a governance-ready reference architecture rather than report statistically generalizable field results. While Quality 4.0 brings in IIoT data, analytics, and AI into quality engineering, many implementations remain tool-focused; models are often integrated into inspection or prediction tasks without being embedded in a PDCA-driven continuous improvement framework, lacking end-to-end evidence preservation and systematic controls for data quality, model drift, and AI risks. This paper presents AITQ-5, a framework governed by PDCA that integrates AI into Total Quality Management for Industry 5.0 manufacturing projects, structured around three design pillars: (i) a PDCA-aligned AI capability model, (ii) human-in-the-loop governance checkpoints throughout the AI lifecycle, and (iii) a Semantic Digital Twin (SDT) that connects Critical To Quality (CTQ) factors, process context, observations, model versions, decisions, and corrective actions into a provenance-aware knowledge graph. The paper offers three implementation artifacts: a PDCA-AI capability map, a maturity model for QMS integration across four levels, and a Quality 5.0 KPI scorecard. Additionally, it outlines an evaluation protocol, case-based sampling guidance, and a roadmap for deployment to support rigorous future validation and audit-ready adoption on the shop floor.