An Edge-AI and Digital Twin Framework for Synchronized Predictive Equipment Maintenance and Automated Supply Chain Reordering

• Asher Noah

  covenant University

  Author

The convergence of artificial intelligence, digital twin technology, and edge computing presents transformative opportunities for U.S. manufacturing ecosystems, yet existing approaches treat predictive maintenance and supply chain reordering as disconnected functions, limiting operational responsiveness and resilience. This study addresses this gap by proposing and validating an integrated framework that synchronizes edge-based predictive maintenance with automated supply chain reordering through a digital twin-enabled decision architecture. Using a mixed-methods design combining retrospective industrial data analysis with prospective simulation, the framework was evaluated across two manufacturing environments: rotating equipment maintenance and multi-line assembly operations. The Long Short-Term Memory (LSTM)-based predictive model achieved 94.8% accuracy in fault detection and 92.6% F1-score, outperforming Random Forest (88.6%) and XGBoost (91.2%) baselines . The edge-AI digital twin implementation reduced latency by 35% (from 125ms to 42ms), decreased cloud bandwidth usage by 28%, and improved fault detection accuracy by 20% compared to cloud-centric architectures . Throughput increased by 18% and process delays decreased by 22% . The synchronized framework enabled 48–72 hour predictive maintenance alerts and automated supply chain triggers, reducing unplanned downtime by 60% and maintenance costs by 28%. This research provides a replicable framework for smart manufacturing practitioners and establishes empirical benchmarks for edge-AI and digital twin integration in industrial operations.

• PDF

Copyright (c) 2026 Asher Noah (Author)

This work is licensed under a Creative Commons Attribution 4.0 International License.