If, in the future, the use of innovative forecasting methods succeeds in reliably predicting the supply of and demand for empty containers per region, this will noticeably improve the basis for planning and decision-making for all stakeholders involved and empty container transports can be reduced to a desirable minimum from an economic and environmental perspective. This is precisely the goal of the C-TIMING research project.
The provision of empty containers at the respective place of loading is a crucial component of global supply chains. According to current estimates, the annual transport costs associated with this alone are in the region of US$20 billion. Against this background, container shipping companies are already making great efforts to reduce empty container shipments as far as possible, both regionally and supraregionally. To facilitate this, the two project partners Container xChange - a logistics marketplace for the brokerage of sea transport containers - and Fraunhofer CML want to create an improved information basis that can support logistics companies, freight forwarders and shipping companies in the efficient and sustainable planning and management of container transports.
Role of the CML in C-TIMING and results:
On the basis of current research results from the field of artificial intelligence, the Fraunhofer CML developed the calculation logic for a container availability index, which predicts the regional and national availability of empty containers. This is based on an evaluation of millions of individual container trips, the implementation of statistical procedures and machine learning techniques to forecast expected supply and demand volumes and their representation using visual analytics methods.
In the course of the project, a unique data pool was built up that reflects the worldwide movements of maritime transport containers. By the end of the project, over 600 million transport events had been recorded and stored in a database. About 30% worldwide container transports could be recorded on a weekly basis at the end of the project. The database was used in the project to develop novel machine learning approaches to forecast future container transports. Based on the forecasts, a key performance indicator, the Container Availability Index (CAI) was developed by the project partners to provide information on regional empty container availability. The CAI can be determined worldwide for individual ports or regions and different container types.
Project Consortium and Funding:
Fraunhofer Center for Maritime Logistics and Services CML, Germany; xChange Solutions GmbH, Germany. C-TIMING was funded by the Federal Ministry of Education and Research from 01.04.2020 – 30.09.2021.