Multi‐objective maritime vessel routing with safety considerations

Abstract The routing of maritime vessels is a challenging optimization problem that involves finding an adequate balance between conflicting and multiple objectives. This paper proposes a methodology based on inverse optimization to find appropriate objective weights that account for conflicting objectives. To formulate the inverse optimization problem, we integrate a weighted multi‐objective function in a model where duality for a network formulation and Karush–Kuhn–Tucker optimality conditions are used as key components. The objective includes route time, fuel consumption, and multiple safety considerations including dynamic stability, the probability of bow slamming, and green water occurrences. The motivation behind our choice of approach lies in the complexity of determining objective weights in multi‐objective problems and the need for incorporating the preferences of multiple stakeholders. To test the proposed approach, we use “best practice routes” based on expert knowledge, real‐world weather data, and domain‐specific objective analysis. Within the scope of this study, these routes are generated using an optimization model with predefined objective weights applied to evaluate the efficacy of the approach. The results demonstrate that the proposed inverse optimization model identifies the weights associated with the best practice routes. A comparison with the analytic hierarchy process (AHP) shows that inverse optimization produces routing decisions more closely aligned with expert‐defined best practice routes, while AHP introduces discrepancies in fuel consumption and travel time, leading to suboptimal routing.