Data-Driven Quality Improvement in Smart Manufacturing: An AI-Enabled Industry 4.0 Framework

Authors

  • Azazullah

Keywords:

Digital twins, Explainable models, Human-in-the-loop decision-making, Model drift, Predictive inspection, Sensor validation, Statistical process control

Abstract

Smart manufacturing increasingly uses artificial intelligence (AI), machine learning (ML), industrial Internet of Things (IIoT), advanced analytics, and digital quality-management practices. Objective: This review synthesizes evidence on data-driven quality improvement and proposes a technology-neutral framework linking data capture, validation, analytical modelling, quality decisions, and continuous monitoring. Methodology: A structured literature review examined research on smart manufacturing, Industry 4.0, AI/ML, manufacturing quality, data quality, Quality 4.0, process monitoring, and optimization using major scholarly databases and supplementary discovery sources. The review used thematic synthesis rather than quantitative meta-analysis. Results: The literature contains many application-specific studies in visual inspection, predictive quality, anomaly detection, and process optimization, but is less consistent in integrating measurement reliability, data governance, model validation, human responsibility, and sustained process control. Recurring barriers include data-quality problems, class imbalance, domain shift, interoperability, explainability, and deployment readiness. The paper proposes a six-stage quality-improvement architecture and a maturity-oriented implementation roadmap. It recommends evaluating technical performance together with process outcomes, operational reliability, and organizational acceptance. Priorities for future research include longitudinal validation, benchmark datasets, continual learning, explainable AI, digital twins, human-centred decision support, and governed use of generative AI.

Published

2026-11-16

Issue

Section

Articles