The encyclopedia · Engineering & Operations · Operational decision · 2021–2024
Xiamen's berth model plans cranes, fairways and safety together
A 0-1 integer program plans berth allocation, crane assignment and fairway traffic at Xiamen's container terminal, lifting throughput.
Xiamen Hai-Tian Container Terminal
the move
Xiamen Hai-Tian Container Terminal at the Port of Xiamen had to allocate berths, assign quay cranes, control fairway traffic and respect berthing safety every day. Treating these separately created plans that were hard to execute safely.
Researchers formulated daily berth planning as a 0-1 integer linear program that integrates berth allocation, quay crane assignment, fairway traffic control and a novel berthing-safety requirement. A highly efficient decomposition heuristic solves the model fast enough for daily use.
The mathematical framework became the core of a berth planning system adopted at the terminal. It transformed the planning process and contributed to increased container throughput and berth capacity.
why it works
- Berth, crane and fairway decisions interact; solving them separately produces plans that conflict.
- Berthing safety is usually left out of models, which is exactly why practitioners distrust them.
- A decomposition heuristic made an integer program practical at daily operational speed.
- A system adopted by planners, not just a paper model, is what changed the terminal's operations.
what transfers
Include the constraints others omit (safety, traffic); a model becomes useful when it matches real operations.
what came after
The berth planning system is in use at Xiamen Hai-Tian Container Terminal and informs further intelligent operations development at the Port of Xiamen and other container ports. The researchers framed the work as a template for mathematical-programming-driven port planning.
references
- Mathematical Programming-Driven Daily Berth Planning in Xiamen Port
- Mathematical Programming-Driven Daily Berth Planning in Xiamen Port
spotted an error? The archive wants to know.