Dynamic barge planning with stochastic container arrivals

Barge transport suffers from a high degree of uncertainty resulting from unreliable container arrivals (e.g. due to deep sea vessel delays), and dynamism that leads to limited information availability during planning. This paper studies the impact of uncertainty and dynamism using real data of an inland terminal. Our method consists of iterating weekly operations for one year to evaluate the long term performance. Each iteration involves solving a stochastic program for planning barge calls, and then simulating actual events. We show that uncertainty has an impact of up to 53% and dynamism up to 20% on total costs.