Navigating complexity in the Great Lakes

Editor’s Note: The SCM thesis Vessel Network Optimization in the Great Lakes Region was authored by Yashar Ahmadov and Sena Perk and supervised by Tim Russell ([email protected]). For more information on the research, please contact the thesis supervisor.

Cargo shipments on the Great Lakes–Seaway system generate $45 billion of economic activity and 238,000 jobs in Canada and the US. However, for shippers, navigating the business environment in this region poses a significant challenge. There are multiple variables and constraints involved in maritime transportation throughout the Great Lakes: Vessels have different speeds and capacities, icing and water levels change over the seasons, trade restrictions vary between the U.S. and Canada, and port capacity limits—all of which come together to make planning difficult.

More vessels, more complexity

Our sponsor company, one of the largest dry bulk shipping services providers in the Great Lakes region, sought to improve the operations of its fleet. The company has a complex network, including U.S. and Canadian ports, and it has recently increased the number of vessels in its fleet. Restricted passage areas, variable drafts, and varying dock and port parameters complicate the network further. Given this complexity, the company wanted to know which trade lanes across the whole market to target with its new combined fleet.

Optimization models help make sense of complexity

We developed a two-stage transportation network optimization model to minimize transportation costs and unloaded vessel sailing time using a quantitative data-driven approach. The solution allocated vessels to trade lanes and then sequenced individual trips. Unfortunately, sequencing problems are among the most computationally challenging problems to solve. Therefore, this study employed clever heuristic approaches and split the problem to reduce the computational burden.

The model consists of two phases: Phase 1, optimal allocation of vessels for each trade lane while minimizing cost; and Phase 2, sequence of trips for each vessel while reducing ballast time. This approach allowed Python code and commercially available optimizers to produce results.

As the firm’s existing planning software did not consider all pertinent operational constraints and planned against current volumes—problematic given the new fleet—our approach was successful. The results show the optimal allocation and sequencing while targeting the whole market. The optimization model allocated and sequenced vessels more efficiently with 20% less ballast ratio per net ton transported.

Every year, approximately 80 students in the MIT Center for Transportation & Logistics’s (MIT CTL) Master of Supply Chain Management (SCM) program complete approximately 45 one-year research projects.

These students are early-career business professionals from multiple countries, with two to 10 years of experience in the industry. Most of the research projects are chosen, sponsored by, and carried out in collaboration with multinational corporations. Joint teams that include MIT SCM students and MIT CTL faculty work on real-world problems. In this series, they summarize a selection of the latest SCM research.

SC
MR