Safety-Oriented Development of Discrete Network Design Problem Considering Accident Index

The network design problem, functioning as a decision-making framework, embodies a policy-based orientation. This problem finds its conceptual footing in the realm of cost optimization, with a key emphasis on the travel time-related expenditures. However, the scope of costs in the network design problem extends beyond travel time. This study undertakes a meticulous exploration of the ancillary costs stemming from safety shortcomings and accidents and introduces the novel concept of the Safe Network Design Problem (SNDP). The investigation of this issue was carried out using the Sioux Falls network as a case study in LeBlanc's model, shedding light on the optimal network improvement strategy. The LeBlanc problem, comprising a selection among five projects within a budget constraint, resulted in the optimal scenario of 10110, indicating the execution of the first, third, and fourth projects. The introduction of safety as a key factor in the problem, particularly in the context of LeBlanc's example, yields intriguing results. In this scenario, the solution shifts to 01110. The combined approach that encompasses travel time costs and accidents, coupled with a comprehensive sensitivity analysis that accounts for accident severity and weighting, engenders a paradigm shift in the solution to the problem.