Abstract
Roads are a crucial component of a country's economic development.Therefore, monitoring and maintaining roads are primary tasks for transportation departments.This study proposes a cost-effective approach to reduce these expenses by introducing a predictive system.This system leverages openly accessible data sources, such as climate and remote sensing information.The foundational framework of this system relies on regression and neural network models, which have been trained using the aforementioned datasets, as well as data from on-site road inspections.These models assign varying weights to input parameters, elucidating the influence of both meteorological conditions and landscape characteristics on the road's overall condition.The resultant system offers forecasts for nine distinct attributes of road sections, each defined by their geographical coordinates.Furthermore, it categorizes the road's condition and provides a visual representation of this data on a geographical map.By exclusively utilizing openly accessible meteorological and remote sensing data, this system has the potential to serve dual purposes, facilitating remote diagnostics and aiding in road design support.