An adaptive large neighborhood search for the multiport continuous berth allocation problem with speed optimization

More than 80% of the cargo is carried by sea, and the global demand is expected to continue increasing in the coming years. This is forcing shipping line carriers and terminal operators to explore new means of optimizing their operations. The berth planning of a terminal is categorized as one of the most critical sea-side operations due to the scarcity of berthing space, and it is modeled as the berth allocation problem (BAP). The aim of the BAP is to assign incoming ships to berthing positions on the terminal, and the continuous version of the problem assumes that ships can berth anywhere along the quay. We extend the continuous BAP to multiple ports, thereby exploiting the potential of a collaboration between carriers and terminal operators. We denominate this problem as the multiport continuous berth allocation problem (MCBAP) and formulate it as a mixed-integer problem. We present an adaptive large neighborhood search (ALNS) heuristic to solve the MCBAP and compare its performance to commercial solvers. ALNS achieves high-quality results within short computational times and presents greater scalability compared to commercial solvers, which makes the ALNS more attractive from an operational and planning perspective.