Automated Scoring Framework for Standardizing the Identification of Icing-Prone Road Sections

Road icing has emerged as a serious threat to traffic safety due to climate change and the increasing frequency of extreme weather events. However, existing evaluations of icy road vulnerability have relied heavily on accident histories or fragmented structural elements and lack standardized and quantitative approaches. To address these limitations, this paper proposes a quantitative scoring system that integrates five domains: climate, terrain, geometry, road facilities, and traffic operation. The framework consists of four major categories and 13 sub-criteria, introducing categorical and graded assessments rather than simple binary judgments, thereby allowing for a more precise evaluation of icing risk. Notably, it distinguishes between highways and local roads, and includes policy applicability as a key factor. A GIS-based empirical analysis was conducted for Seoul. The results identified highly vulnerable zones along major elevated roadways, such as Gangbyeonbuk-ro, Olympic-daero, and Dongbu Expressway, which correspond to actual accident patterns. In addition, factors often overlooked in previous frameworks, such as insufficient sunlight duration, high north-facing slope ratios, and imbalanced distribution of de-icing infrastructure, were effectively captured. This study demonstrates that the proposed framework can shift icy road management from reactive response to proactive prediction and prevention, thereby providing a foundation for applications in smart de-icing strategies, GIS-based hazard mapping, AI-driven risk forecasting, and the development of national standardized guidelines.