Road classification in Chiang Mai City using Link and Place concept

Abstract

Urban road classification is a critical mechanism in planning the development of traffic infrastructure and land use management efficiently. However, conventional road classification systems that focus solely on mobility—such as the Functional Classification System (FCS) and the Expanded Functional Classification System (Expanded FCS)—are limited in their ability to reflect the evolving land use patterns and community activities in urban areas. This article presents the application of the Link and Place framework to classify roads within the Second Ring Road area of Chiang Mai Province, covering a total length of 217.58 kilometers. The Link dimension was assessed based on physical infrastructure and Average Annual Daily Traffic (AADT), while the Place dimension was evaluated through land use patterns interpreted from satellite imagery using U-net convolutional neural networks, analyzed on a 300×300-meter grid. The results showed that Link 5 roads (community-level roads) accounted for the largest proportion of the network at 43%, followed by Link 4 (24%) and Link 3 (22%). For the Place dimension, 54% of roads were located in high-density suburban areas and 24% in dense urban areas. The findings indicate that the traditional classification system fails to capture the spatial relationship between travel function and land-based activity, as some roads with intensive land use and community activities were still classified as primary roads. The application of a Link × Place matrix allows for a more context-sensitive approach to street network planning and development, addressing both mobility and quality of urban life. This framework serves as a valuable tool for sustainable urban planning and future infrastructure design.