Human-Centric automation to intelligent information integration: A mixed-methods framework for industry 5.0 manufacturing

This study examines how human expertise, data governance maturity, and AI-driven analytics jointly influence decision quality in manufacturing environments transitioning toward Industry 5.0. Using a mixed-methods design, a structural equation model (n = 82) and conducted a thematic analysis of 398 coded statements from survey and interview data was developed and tested. Quantitative results show that AI-driven analytics has a strong positive effect on decision quality, whereas human expertise and data governance exhibit negative associations when operating within fragmented or weakly integrated information systems. Qualitative analysis explains these patterns by revealing structural barriers including inconsistent data semantics, metadata loss, legacy machinery, and manual data-handling that diminish the reliability of intelligent decision support. To synthesise these findings, the Human–AI Information Integration Framework is proposed, which conceptualises decision quality as an emergent property of hybrid intelligence. The framework demonstrates how human contextual reasoning, robust governance structures, and AI-enabled analytical capabilities interact to support adaptive and trustworthy decision-making. This research advances the intelligent systems literature by empirically linking socio-technical theory with operational AI deployment and by identifying the conditions under which intelligent information systems enhance rather than destabilise organisational decision processes. The findings provide practical guidance for designing human-aligned, interoperable, and resilient intelligent manufacturing systems.