Smart Infrastructure Maintenance: A Review of City Infrastructure Maintenance Technology and Methods for Decision Support System

Infrastructure assets such as roads, drainage systems, and bridges are important components of a city. Research initiatives have focused on streamlining various infrastructure project activities, particularly in the maintenance phase. This has been achieved through the implementation of diverse digital methods and technologies, which automate process at each stage. This research aims to offer a deeper understanding of existing studies and to establish a foundation for future research initiatives. To achieve this, a comprehensive review was conducted on the maintenance processes of infrastructure, emphasizing technological and automation methods. The analysis focused on 40 published articles, exploring various facets such as location and mapping, inspection and monitoring, condition assessment, and priority decision-making and improvements. Through this review, significant limitations and challenges were identified, including the constraints of traditional image processing, the time and effort required for deep learning model development, reliance on historical data, inconsistencies in condition ratings, and the lack of predictive maintenance approaches. Subsequently, a status matrix was developed to evaluate the technological methods examined and their potential for future research. Ultimately, this study highlights emerging trends for future exploration, specifically in the areas digital twin, quality and uncertainty of inspection data, data generation and semi-supervised learning, and predictive maintenance. By providing a systematic review of the technologies and methods applied in infrastructure maintenance activities, this research not only identifies existing limitations and challenges but also highlights key research trends, paving the way for future investigations.