IPIC 2026 webinar III Artificial Intelligence Applications for the Physical Internet

The third webinar in the IPIC 2026 series, Artificial Intelligence Applications for the Physical Internet, explored how artificial intelligence (AI) can support the transition towards more collaborative, efficient and zero-emission logistics systems. 

Organised in collaboration with ALICE, the EU-funded IKIGAI project and KEDGE Business School, the session focused on how data-driven decision-making and optimisation methods can enable the operationalisation of the Physical Internet (PI), particularly in urban logistics. 

Opening the session, Pablo Segura highlighted the role of AI within ALICE’s vision of climate-neutral and competitive freight transport. As a European Technology Platform, ALICE connects stakeholders across the logistics ecosystem to accelerate innovation through collaboration. Within this framework, the Physical Internet enables interconnected and shared logistics networks, while AI provides the intelligence required to manage complexity, coordinate flows and support real-time decision-making. 

The webinar also reflected IKIGAI’s mission to accelerate the twin transition towards green and digital logistics through industry-led innovation. By demonstrating practical use cases, IKIGAI contributes to making Physical Internet concepts operational and scalable. 

From open data to shared logistics maps 

In the first presentation, Tianyuan Zhang addressed the lack of a shared spatial foundation in urban logistics. Currently, operators define delivery zones independently, while cities rely on administrative boundaries that are not suited to logistics operations. This fragmentation limits collaboration and complicates both planning and regulation. 

The proposed framework, developed within the SOUL project, introduces a unified approach to urban logistics districting based on open data. By combining road network data and building information, the method creates a multi-tier structure of spatial units, ranging from small base units to larger operational districts. These are generated through clustering techniques and optimisation models, ensuring both scalability and reproducibility across different cities. 

Applied to diverse urban contexts such as Bordeaux, Manhattan and Chengdu, the framework demonstrated strong adaptability. From a Physical Internet perspective, it provides a shared operational map that enables alignment between stakeholders, supports benchmarking and strengthens the basis for policy design. 

AI-driven adaptive collaboration 

The second presentation by Nafe Moradkhani focused on how AI can enable adaptive collaboration in logistics networks. While the Physical Internet promotes openness and resource sharing, the presentation emphasised that collaboration must be context dependent. Differences in operational constraints, demand patterns and local conditions mean that full collaboration is not always optimal. 

The proposed approach combines machine learning with optimisation to determine when and how collaboration should occur. A predictive model evaluates contextual signals such as workload, congestion and service performance to estimate the appropriate level of openness. Based on this, optimisation models define operational decisions, including resource allocation and parcel transfers between actors. 

The findings show that selective collaboration can significantly improve performance without requiring full system integration. Moderate levels of openness were sufficient to reduce overload and enhance resource utilisation, highlighting that adaptive and context-aware collaboration is a practical pathway towards implementing Physical Internet principles. 

Conclusion 

The webinar demonstrated that AI plays a key role in enabling the Physical Internet by supporting shared infrastructures, improving coordination and enabling more transparent, data-informed decision-making. These capabilities are essential to scaling collaborative logistics solutions and achieving zero-emission freight transport. 

As IKIGAI continues to showcase innovation in this field, AI-driven approaches will remain central to advancing sustainable and interconnected logistics systems. The discussion will continue at IPIC 2026, where stakeholders will further explore the future of Physical Internet deployment.