AI Driven Predictive Maintenance at the Edge

Abstract

Predictive maintenance (PdM) has emerged as a transformative strategy in industrial and manufacturing domains, allowing organizations to anticipate equipment failures, reduce downtime, and optimize operational costs. The integration of Artificial Intelligence (AI) with edge computing has enabled real-time analytics, rapid decision-making, and localized processing of large-scale sensor data. This review paper discusses the architecture, methodologies, and applications of AI-driven predictive maintenance at the edge, highlighting its benefits, limitations, and future prospects. The paper also presents case studies and illustrative models for deploying AI at edge nodes to enable proactive maintenance strategies in diverse industrial environments.