A Multi-Objective Optimization Model to Minimize the Gap in Accessibility to Multi- Use Paths While Maximizing the Economic Efficiency of Active Transportation Investments for Fresno, California

This research examines the accessibility to multi-use paths in a variety of neighborhoods in Fresno, California. This study aims to develop a multi-objective optimization modeling framework to be used by decision makers in transportation and policy to maximize the total accessibility to multi-use paths across the city while minimizing the gap between low- and high- accessibility neighborhoods by an optimal allocation of active transportation. Researchers calculate accessibility to multi-use paths by measuring the total length of multi-use paths (walkway and bikeway) a resident could reach from their own Census block group with a 30-minute cycling ride. The study employs a geographically weighted regression (GWR) model to capture the local relationships between accessibility to multi-use paths and previous transportation investments (walkways, bikeways, and primary and secondary roads), while controlling for other socioeconomic factors. The marginal-effect analysis for the GWR results categorizes the areas within Fresno into economically efficient, inefficient, and indifferent locations for further investments. The researchers embed the GWR results into a multi-objective optimization modeling framework to improve accessibility to multi-use paths over the city and simultaneously address inequality in active-transportation accessibility. This research provides decision makers with insight into the problem of making of an economically-efficient and socially-equal active transportation plan accessible to people from diverse backgrounds with the ultimate goal of fostering public health.