As part of deliverable 2.3 – ReMuNet machine processable objective functions, PTV Group has made a substantial contribution by developing machine-processable objective functions for optimised multimodal and synchromodal freight transport. These functions are designed to support automated, constraint-aware decision-making across complex transport and logistics networks, enabling more resilient, efficient, and environmentally sustainable freight operations.

At the heart of this work there is a mathematical framework that transforms transport planning into a solvable optimisation problem. The models encode diverse operational requirements—including cost, emissions, time windows, and capacity constraints—into graph-based structures that can be processed using linear and integer programming techniques. This approach enables planners to allocate resources intelligently across multiple transport orders and intermodal routes.

Advanced modelling for complex logistics scenarios: Building on PTV’s intermodal routing engine, Deliverable 2.3 introduces a key advancement: the ability to optimise multiple freight orders in parallel. This is achieved through two modelling approaches. The first—computationally efficient—assumes a single intermodal leg per order, while the second, based on the Multicommodity Network Flow Problem (MNFP), supports complex, multi-leg intermodal chains. The latter offers greater realism and flexibility, especially for integrated planning across shared infrastructure.

Each transport option is treated as a distinct “resource” within a directed graph, defined by location, timing, cost, and capacity. These are then optimised using a Gurobi solver, integrated via a server-based API. To ensure practical usability, PTV developed a user interface that allows planners to configure constraints, evaluate scenarios, and interpret optimisation outcomes with key performance indicators such as cost, time, and CO₂ emissions.

PTV’s work in Deliverable 2.3 leads to two key conclusions—each underscoring the real-world value and future potential of the developed solution.

Conclusion 1: Supporting—but not replacing—the human planner

Freight planning is a demanding task that requires continuous attention and adaptation. Throughout the day, planners must respond to disruptions, shifting capacities, and changing customer demands. The optimisation system developed by PTV acts as an intelligent assistant, delivering semi-automated support to answer a central question: Which transport chain should be selected for a given order, under current conditions?

By constantly recalculating optimal solutions in response to new data—while respecting constraints related to time, capacity, emissions, and cost—the system helps planners make faster, better-informed decisions. Crucially, the planner remains in full control. Their contextual knowledge and experience remain indispensable, particularly in scenarios where qualitative factors cannot yet be captured by the model.

Conclusion 2: A scalable, API-Based solution for collaborative use

The second conclusion focuses on the system’s architecture and scalability. Designed as an API-based solution, the optimisation service can be flexibly deployed both within ReMuNet and in wider collaborative platforms. In future use cases, multiple logistics providers could jointly optimise transport orders and infrastructure usage—unlocking significant gains in efficiency and sustainability.

This collaborative optimisation capability opens the door to achieving a global system optimum—a state where resources are utilised as effectively as possible across actors, with benefits for cost, emissions, and service reliability.

Following successful validation with test data and dynamic scenarios, the solution will be further refined in Task 2.4 and upcoming pilot activities using real-world data. These next steps will assess integration into operational workflows and further align the system with planners’ day-to-day needs.

PTV’s work for this deliverable represents a strong step toward intelligent, integrated freight planning, bringing automation, optimisation, and human expertise together to meet the demands of a dynamic transport landscape.

Read the deliverable here.