Optimization of Buffer Times and Recovery Actions in Liner Shipping Networks



The main goal of this paper is to develop a method to help liner shipping networks to cope with delay and its costs. To do so, the paper studies two types of policies to prevent and recover delay, firstly, to assign buffer times to different stages of a trip in the scheduling and secondly, to perform recovery actions during the trip in case delay occurs. A mathematical optimization model is formulated to determine how these policies should be implemented in order to minimize the total cost of the trip including delay cost and cost of recovery actions. The model formulates delay as a stochastic phenomenon depending on exogenous factors, recovery actions and buffer times. To solve the model, the paper develops a two-phase global optimization algorithm based on stochastic dynamic programming. The first phase tries to find a good feasible solution for the problem and the second phase finds the optimal solution by a branch and bound algorithm. Finally, the algorithm is tested on a real problem and finds its optimal solution in a short time.
The Seminars in Econometrics Series is supported by the Tinbergen Institute, ERIM and the Journal of Applied Econometrics.  This extra seminar is also supported by the Erasmus Research Centre on Business Intelligence (ECBI).
Contact information:
Remy Spliet                     
Email kbouwman@ese.eur.nl