Tugboat scheduling under ship arrival and tugging process time uncertainty

• Published in: Transportation research. Part E, Logistics and transportation review Vol. 144; p. 102125
• Main Authors: Kang, Liujiang, Meng, Qiang, Tan, Kok Choon
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.12.2020
• Subjects: Container port; Proactive scheduling strategy; Reactive scheduling strategy; Tugboat scheduling; Uncertainty; Uncertainty; Reactive scheduling strategy; Container port; Proactive scheduling strategy; Tugboat scheduling
• ISSN: 1366-5545; 1878-5794
• DOI: 10.1016/j.tre.2020.102125

•We consider the tugboat scheduling problem under ship arrival and tugging process time uncertainty.•The uncertain ship arrival and tugging process times are modeled by a finite set of discrete scenarios.•The proactive scheduling strategy and reactive scheduling strategy are both considered in the problem.•A MILP model is built to minimize the total weighted service time and the expected value of recovery cost.•An ad-hoc algorithm is designed to generate tugging chains such that the large-scale problems can be tackled effectively. This study addresses an interesting tugboat scheduling problem considering uncertainty in both container ship arrival and tugging process times for large container ports. The uncertain ship arrival and tugging process times are formulated as a finite set of discrete scenarios that can be generated from historical port traffic data. We deal with the uncertainty by integrating proactive and reactive scheduling strategies such that this study is distinct from most existing studies in the literature. The proactive scheduling strategy considers the expected degree of variability and uncertainty during the execution of a tugboat fleet schedule while the reactive scheduling strategy properly adjusts the initial schedule to cope with unexpected scenarios with minimum recovery cost. A mixed-integer linear programming model for the proposed tugboat scheduling problem is established. For a large-scale problem, an ad-hoc algorithm is designed to generate tugging chains such that the large-scale problem can be tackled effectively. The extensive numerical experiments are finally carried out to demonstrate the practical significances of the models and algorithms developed by this study.