Home > Logistics and distribution > Optimnet project

Optimnet project

Success story

France’s national postal service, La Poste, was facing rapidly-increasing demand for parcel delivery services. The organization was looking for a way to transform its delivery route design system, a large and complex ecosystem made up of many interconnected national and regional lines. La Poste turned to Probayes for support with the transformation, which came in the form of a custom decision-assistance tool used to test scenarios and explore new ways to improve the network.

Challenge

Each problem is broken down into three steps:

The future parcel delivery network is modeled
The routing solution is optimized by determining the specific route for each parcel (stopovers, routes), the logistics operations completed (sorting, cabotage), the types of vehicles used, and the delivery time
The tool can be used to complete simulations on a regular basis to test new scenarios, based on predicted volumes, the characteristics of the logistics chain (sites, interconnections), and economic hypotheses.

The project

Probayes’ operational research team worked on all three of these steps:

They modeled the parcel distribution network using a linear mathematical model tailored to the use case being addressed and integrating the different possible improvements to the network, which were determined with experts from La Poste. The models can have more than 1 million variables and more than 20,000 constraints, depending on the number of sites involved and the routing strategies considered.
A program was implemented in C++ for the construction of mixed linear programs, solved with a commercial solver, and the reconstruction of the solutions
A simulator was developed for users to configure scenarios, run simulations, display the optimized solutions, and compare the results based on a number of indicators. The web-based graphical user interface was developed in Python

Results

The tool produces optimized solutions (within 5% of optimal) in less than ten minutes. The average transportation-related savings at the national level are between 5% and 20% per day. Initial testing of new scenarios generated several promising potential improvements to the network. The simulator is now available to the Colissimo parcel delivery service division of La Poste for testing.

A funded PhD dissertation in cooperation with G-SCOP Lab should result in enhancements to the resolution methods used. The tool will then be able to generate even more effective solutions and handle an even greater number of variables, still within a reasonable amount of time.

“Lorem ipsum dolor sit amet, consectetur adipiscing elit. Fusce eu lacus diam. Cras congue, neque ac mattis efficitur, nibh ipsum fringilla nunc, sed iaculis neque odio non orci. Vestibulum ante ipsum primis in faucibus orci luctus et ultrices posuere cubilia Curae; Praesent lacinia, urna quis rhoncus lobortis, urna neque tempus tellus, quis auctor justo lectus vitae libero. Nulla non porta odio. Donec diam est, varius id ullamcorper a, efficitur nec libero”