Generation of Container Flows for Container Terminals
In maritime logistics, simulation and mathematical optimization are commonly used methods for planning, solving problems and evaluating solutions. For testing new solutions under realistic conditions, an extensive amount of actual operational data are urgently needed. Since comprehensive real-life data are often unavailable or classified, the generation of synthetic data is a helpful way to get around this issue. One such instance is when we want to study of the dynamics at a container terminal, the nodes which connect the seaside with the hinterland, transshipping containers between vessels of different sizes, trains, and trucks. The container terminal functions as a buffer between the modes and offers to keep the containers in the yard for some hours or days. The vehicle arrivals and the containers to be unloaded and loaded from the vehicle constitute the container flows through a terminal.
With ConFlowGen, short for Container Flow Generator, an IT tool has been developed at the Institute of Maritime Logistics which is capable of generating such synthetic scenarios. It has been first presented at the LDIC conference 2022 and has been subject to continous development since then. The tool has been deployed in internal projects at the institute as well as in theses of students. The source code is hosted at https://github.com/1kastner/conflowgen and is open source. The project is supervised by Marvin Kastner.
In case the project has sparked your interest, we'd like to learn more about you and maybe even your use case!
| [192106] |
| Title: Addressing Challenges in Creating Traffic Profiles for Transshipment Hubs in Seaports. <em>Logistics and Maritime Systems: 12th International Conference on Logistics and Maritime Systems (LOGMS 2024) - Proceedings</em> |
| Written by: Gupta, Shubhangi and Kastner, Marvin and Vieira, João and Jahn, Carlos |
| in: (2024). |
| Volume: Number: |
| on pages: [IN PRINT] |
| Chapter: |
| Editor: In Voß, Stefan and Heilig, Leonard (Eds.) |
| Publisher: Springer: |
| Series: Lecture Notes in Computer Science |
| Address: Germany |
| Edition: |
| ISBN: |
| how published: |
| Organization: |
| School: |
| Institution: |
| Type: |
| DOI: 10.13140/RG.2.2.34646.51520 |
| URL: https://easychair.org/publications/preprint/glmV |
| ARXIVID: |
| PMID: |
Note: conflowgen
Abstract: Globalization has significantly increased containerized traffic, driven by the rising demand for swift cargo movements at low cost. When creating cost- and time-efficient maritime transport networks, transshipment hubs are of high importance. There, containers are moved from one vessel to another. This enables carriers to design hub-and-spoke networks(where feeder vessels serve the spokes and deep sea vessels interconnect the hubs) as well as connecting deep sea services by interlining. Hubs are often located along major shipping routes and are concentrated near canals and straits. The successful operation of transshipment hubs relies on various socio-economic factors, trade policies, and robust infrastructure. When making strategic decisions, simulation is often used to estimate the impact of each viable option on terminal performance. Such simulation studies heavily depend on suitable synthetic traffic profiles that reflect the workload and yard occupancy of transshipment hubs over a longer time horizon. Past work has shown that for transshipment hubs, the expected average yard occupancy is approximated over the course of several weeks, which increases the runtime of simulation studies. The approach presented in this paper addresses this issue by modifying the code of library ConFlowGen and applying it on three use cases. The results show that the traffic profiles generated with the modified code are more suitable for simulating operations of transshipment hubs. Several traffic profile characteristics are discussed, that are difficult to satisfy at the same time.