Digital Twin-Enabled Intelligent Computing Frameworks for Smart Manufacturing and Predictive System Maintenance: A Comprehensive Review

Abhishek Banerjee, Pooja Gupta, Parul Joshi

Abstract


ABSTRACT:  In the era of Industry 4.0 and smart manufacturing, the integration of Digital Twin (DT) technology with intelligent computing frameworks has emerged as a cornerstone for enhancing operational efficiency, minimizing unscheduled equipment downtime, and enabling proactive predictive maintenance (PdM). Digital Twins create dynamic, high-fidelity virtual representations of physical assets, manufacturing lines, and complex industrial processes through continuous real-time synchronization with Industrial Internet of Things (IIoT) edge nodes. However, processing massive, heterogeneous, high-frequency sensor telemetry streams in real time poses unprecedented challenges regarding computational latency, storage complexity, multi-source data fusion, and predictive accuracy. This comprehensive review paper systematically explores state-of-the-art Digital Twin-enabled intelligent computing frameworks tailored for smart manufacturing environments and predictive system maintenance. We evaluate architectural paradigms spanning edge, cloud, and hybrid edge-cloud computing configurations, along with advanced data fusion methodologies and artificial intelligence algorithms including Physics-Informed Neural Networks (PINNs), Deep Reinforcement Learning (DRL), dynamic Transformers, and automated Remaining Useful Life (RUL) estimation models. Furthermore, this study presents a rigorous taxonomy of maintenance paradigms, provides benchmark experimental comparisons, addresses critical technical bottlenecks, and articulates future research directions to pave the way toward self-healing, fully autonomous cyber-physical manufacturing systems.


KEYWORDS: Digital Twin, Smart Manufacturing, Intelligent Computing, Predictive Maintenance (PdM), Remaining Useful Life (RUL), Edge-Cloud Architecture, Physics-Informed Neural Networks.


Full Text:

PDF 79-90

Refbacks

  • There are currently no refbacks.


Copyright 2017 ManTech Publications. All Rights Reserved