AI-Based Predictive Maintenance in Computer Engineering: Enhancing System Reliability Under the Moderating Role of Edge Computing Integration

Abstract

In the era of digital transformation, ensuring high system reliability in computer engineering has become a critical priority, particularly in distributed and resource-constrained environments. Artificial Intelligence (AI)-based predictive maintenance, and specifically Failure Pattern Detection via Machine Learning (FPDML), offers a proactive approach to fault management by identifying anomalies before they cause system downtime. However, the effectiveness of these strategies is influenced by real-time data processing capabilities, with Edge Computing Integration (ECI) emerging as a key enabler to reduce latency, enhance fault tolerance, and strengthen operational continuity. This study employs a quantitative research design to examine the impact of AI-based predictive maintenance on system reliability, with FPDML as a core mechanism and ECI as a moderating variable. Drawing on Reliability Theory, Edge AI, and Distributed System Reliability frameworks, the research analyzes responses across Vietnam, Singapore, and Malaysia, including system engineers, IT administrators, operations managers, and policy researchers. Data was collected using a structured 5-point Likert-scale survey and analyzed to assess direct and moderated relationships among variables. The findings are expected to provide empirical evidence on how AI-driven maintenance strategies, when integrated with edge computing, can significantly improve system uptime, reduce unplanned outages, and optimize resource usage in complex computing environments. The study also addresses practical challenges such as model interpretability, data quality, and cybersecurity, offering insights for both industry practitioners and policymakers aiming to enhance the reliability of next-generation distributed systems.