INTEGRATING SHORTEST-PATH ANALYSIS and MULTI-AGENT SIMULATION for CAMPUS ACCESSIBILITY EVALUATION

Physical accessibility constitutes a fundamental dimension of spatial justice, ensuring that all individuals can equally benefit from the built environment. This study aims to evaluate and quantify the accessibility challenges faced by wheelchair users within the Sivas Cumhuriyet University campus and to propose a model that objectively identifies spatial inequalities. For this purpose, the intra-campus pedestrian network was modelled using Python, and shortest-path analyses were conducted via Dijkstra’s algorithm for 1,000 randomly selected origin–destination pairs. These routes were further simulated within a multi-agent system (MAS) environment to estimate travel times and compare mobility performance between wheelchair users and able-bodied individuals. This study provides one of the first quantitative frameworks to integrate Dijkstra-based shortest-path computation and MAS-driven simulation for assessing wheelchair accessibility in outdoor environments. The resulting data were used to develop a numerical accessibility scoring system that expresses spatial disadvantage as an accessibility coefficient. The findings revealed that 85.8% of the routes were completely inaccessible for wheelchair users and that, where access was possible, travel distances were on average 8.5 times longer than those of non-disabled individuals. By establishing a reproducible and data-driven framework, the study connects the aim of promoting spatial equity with quantifiable outcomes, thereby providing a decision-support tool for campus redesign and urban accessibility planning. These findings provide a scalable analytical framework for promoting spatial equity, offering practical guidance for policymakers and urban planners seeking to improve accessibility in built environments.