Edge AI, Industrial IoT, and Autonomous Control for Sustainable Predictive Manufacturing: A Comparative Engineering Systems Analysis of the United States and Taiwan

Abstract

This article investigates how edge artificial intelligence, Industrial Internet of Things architectures, and autonomous control systems influence predictive manufacturing, operational optimization, engineering resilience, and sustainability outcomes in advanced industrial ecosystems. Using a comparative engineering systems analysis of the United States and Taiwan, the study examines two technologically advanced yet structurally different industrial cases: the United States as a large-scale, diversified, AI-intensive manufacturing and digital infrastructure ecosystem, and Taiwan as a highly specialized, semiconductor-centered, precision manufacturing system with dense cyber-physical integration. The study draws on engineering literature, industrial automation indicators, OECD digital transformation evidence, International Energy Agency energy-efficiency data, International Federation of Robotics reports, World Economic Forum smart manufacturing analyses, and IEEE-indexed studies on edge computing, industrial IoT, predictive maintenance, and autonomous manufacturing control. The findings demonstrate that edge AI reduces decision latency, improves predictive maintenance responsiveness, enhances quality-control precision, and supports energy-aware production when integrated with robust industrial data infrastructure. However, the comparison reveals that technological performance depends on system architecture, semiconductor supply-chain integration, interoperability standards, cybersecurity governance, workforce capability, and sustainability alignment. This article contributes to engineering and applied science scholarship by proposing an integrated framework linking sensorized manufacturing infrastructure, edge AI analytics, autonomous control, operational optimization, engineering resilience, and sustainable socio-technical industrial development.