A MCDM Based Approach to Prioritizing National Highways for Road Safety Improvements

Selection of the most vulnerable alternative for implementation of road safety projects considering financial and technical availability often put the road authorities in dilemma. This paper proposes a two-step approach based on multi-criteria decision-making methods to overcome the difficulties in location prioritization. The objective of this study was to rank the national highways based on their vulnerability in terms of road safety and identify the location that requires the most attention. The study area covered Cumilla-Feni section of N1 national highway, Gazipur-Elenga section of N4 national highway, Natore-Nawabganj section of N8 national highway, and Barisal-Madaripur section of N8 national highway. These four alternatives were evaluated under five criteria- average annual daily traffic (AADT), crash per thousand vehicles, percentage of corridor without median, heavy vehicles percentage in the corridor, and percentage of non-motorized vehicles in the corridor. The required data for analysis were collected from some secondary sources along with drive-through video footage from a probe vehicle. To evaluate the relative importance of each criterion, criteria weights were calculated using the Analytic Hierarchy Process (AHP). AHP uses pair-wise comparison based on subjective judgment to establish relative importance and the consistency ratio provides a measure of the consistency of the judgment. Here, the consistency ratio was found 0.02 which does not exceed CR = 0.1, indicative of consistent judgment. The criteria weights and the criteria values were combined using the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) and ranks of alternatives were achieved. The analysis identified Natore-Nawabganj (N6) as the most vulnerable alternative followed by Gazipur-Elenga (N4), Cumilla-Feni (N1), and Barisal-Madaripur (N8). This study proposes a framework for evaluation of alternatives that is reliable, and data driven. The results indicate that the proposed framework possesses replicability and promises effective decision-making based on scientific approach.