Use Case #5: Smarter Inland Shipping and Container Transport

We will develop AI-driven tools to improve the efficiency, speed, and flexibility of inland shipping and container transport planning.

Background

Container transport to the hinterland is under increasing pressure, as road, rail, and inland waterway networks are operating close to their capacity limits. At the same time, upcoming large-scale infrastructure maintenance and growing transport demand will further strain the system. This makes it increasingly difficult to plan efficient and reliable transport flows.

Current planning approaches are often fragmented and focus on optimising individual shipments or objectives, which can lead to inefficiencies at the system level. Although planning tools already exist, they are not always used in practice, partly due to limited trust, lack of transparency, and a mismatch with how planners make decisions.

At the same time, the availability of data and advances in AI create new opportunities to better coordinate transport across modalities—road, rail, and inland waterways—and to move towards more integrated and flexible planning.

Objective

The objective of this use case is to develop and test AI-driven planning tools that support more efficient and flexible container transport across different modalities.

The focus is on improving the utilisation of available capacity—particularly in inland shipping—by enabling better coordination of transport flows at the system level. This includes balancing multiple objectives such as efficiency, speed, and societal impact.

In addition, the use case aims to understand how AI-based planning tools can be effectively used in practice. This includes addressing barriers to adoption, increasing user trust, and ensuring that AI-generated insights can be translated into actionable decisions for planners.

Approach

The use case combines AI development with practical validation and user involvement. AI-based optimisation models—building on techniques such as reinforcement learning—are developed to handle complex, multi-variable planning problems and to evaluate alternative transport scenarios in real time.

These models are developed and tested in close collaboration with end users, including transport planners, to ensure that the tools align with real-world decision-making processes. Particular attention is given to explainability and usability, so that planners can understand and trust the recommendations provided by the system.

In parallel, different planning approaches are compared to assess the added value of AI-based optimisation over existing methods. The results are translated into practical tools, guidelines, and learning modules, contributing to both operational improvements and knowledge development within the wider Learning Community.