Abstract: Machine learning (ML)–driven frameworks for intelligent decision-making, predictive analytics, and adaptive control in cyber–physical production systems have propelled smart manufacturing. The complexity of automated and robotic production contexts makes ML model transparency and explainability a major research topic. Recently developed deep learning–based defect prediction methods may capture complicated temporal degradation patterns from high-dimensional sensor data, but their interpretability limits industrial use. In light of these findings, this theoretical exploration examines ML-driven frameworks for intelligent and explainable manufacturing systems using predictive maintenance, Natural Language Processing (NLP), and Large Language Models (LLM) approaches. The study aims to conceptually investigate how deep learning–based time-to-fault prediction methods might be used to explainable manufacturing frameworks for human-centric decision-making. The qualitative study synthesizes literature and references deep learning architectures like sequence learning and temporal modelling utilized for time-to-fault prediction in automated manufacturing and humanoid robotics. Theory-mapped to intelligent manufacturing systems, these techniques use explainability mechanisms, NLP-based semantic interpretation, and LLM-assisted knowledge extraction to turn model outputs into human-understandable insights. Theoretical studies imply that ML-driven predictive frameworks can reduce unplanned downtime, improve maintenance planning accuracy, and improve system dependability, while NLP and LLM components help operators make contextual decisions. The study concludes that transparent and trustworthy intelligent manufacturing systems require explainable ML, NLP, and LLM technologies. The consequences include greater operational efficiency, trust in AI-driven production decisions, and a scalable theoretical underpinning for empirical research and Industry 5.0 implementations.
Keywords: Machine Learning–Driven Frameworks; Explainable Manufacturing Systems; Intelligent Manufacturing Systems; Machine Learning; Natural Language Processing; Large Language Models.
Title: A Study of Machine Learning–Driven Frameworks for Intelligent and Explainable Manufacturing Systems: A Theoretical Exploration
Author: Suraj Shrestha
International Journal of Computer Science and Information Technology Research
ISSN 2348-1196 (print), ISSN 2348-120X (online)
Vol. 14, Issue 1, January 2026 - March 2026
Page No: 128-134
Research Publish Journals
Website: www.researchpublish.com
Published Date: 09-March-2026
