Deep Learning for Predictive Maintenance in Critical Infrastructure: A Study on Smart Grid Systems

This study synthesizes recent advancements in artificial intelligence (AI), Internet of Things (IoT), Industry 4.0/5.0, and immersive computing technologies that drive sustainable innovation across sectors such as urban mobility, energy management, manufacturing, and smart cities. Emphasizing the critical role of AI-enabled predictive maintenance in smart grid infrastructure, the research investigates the accuracy of deep learning models in forecasting failures and their impact on operational efficiency, downtime reduction, and cost savings. Utilizing a quantitative methodology with a representative sample of 480 respondents from Gujarat, data were collected via structured questionnaires employing a five-point Likert scale and analyzed through multiple regression models in R Studio. Findings reveal that machine learning accuracy significantly enhances predictive maintenance effectiveness, while failure rate reduction demonstrates a nuanced influence. The study underscores the transformative potential of digital twins, cyber-physical systems, and human-centric Industry 5.0 frameworks in promoting resilient, sustainable urban and industrial ecosystems. Future research should integrate AI-driven predictive analytics with digital twin technologies and explore socio-economic dimensions of technology adoption to ensure equitable and scalable smart infrastructure development. This research contributes to bridging technological innovation with sustainability goals, informing policymakers and industry stakeholders in the global transition toward smart, efficient, and inclusive infrastructures.