The LabCom LOPF,
a unique research
structure

A unique model of research cooperation: the LabCom LOPF (Large-scale Optimization of Product Flows), a team of world-renowned experts in partnership with leading academic institutions:

  • CNRS
  • Sorbonne Université
  • Laboratoire de Probabilités Statistiques et Modélisation (LPSM)
  • Université Paris Cité
  • SCAI (Sorbonne Center for Artificial Intelligence)

We design tomorrow's solutions for optimizing large-scale flows of fresh produce in an environmentally friendly and socially responsible way.

This technology is deployed within the market of Rungis.

Califrais lab

Logistics 4.0

Califrais has won the "Logistique 4.0" call organized by the ADEME as part of the "France 2030" plan. The aim of this project is to take the solutions developed by Califrais in Rungis to an industrialization phase across the entire network of French wholesale markets, in order to leverage the benefits and impact of the technologies being developed.

The 22 French food wholesale markets are key multimodal hubs for national food security, resilience and sovereignty, as well as for the organization of urban logistics. However, the logistics flows between these markets, as well as their economic and environmental costs, remain largely under-optimized as they are neither mutualized nor centralized.

Focus on some research topics

Our research topics are diverse, at the intersection of machine learning, logistics optimization and ecology. Within the Labcom, our PhD students work at the interface between academia and industry on topics such as:

  • Online convex optimization algorithms for large-scale inventory problems
  • Adaptive multi-horizon probabilistic time-series forecasting to model daily demand for a catalog of thousands of fresh produce items
  • Optimization of operational (logistics and transport) and ecological costs, taking into account the specific constraints of our sector
  • Survival models with high-dimensional longitudinal data for modeling customer satisfaction
  • Modeling the dynamic interactions of agents in wholesale markets

Time series analysis

Reinforcement learning

High-dimensional statistics

Online learning

Natural language processing

Deep learning

Survival analysis

Statistical learning