Positioning cyclo-logistics depots in Bordeaux
Work state summary
Maximum number of files: 10
Type of work: Individual work
With more people and more online shopping, cities are experiencing an explosion in the number of parcels. In 2017, 505 million parcels were delivered in France, and by 2021, retailing had reached €129 billion. More parcels means more trucks and more pollution. The last kilometre of delivery is the worst, because it's costly and highly polluting. Add to that the fact that 31% of greenhouse gas emissions in 2018 came from transport, and you have an idea of the puzzle!
What if we did it all by bike? Cyclo-logistics is the art of delivering goods by bike. Imagine cargo bikes, three-wheelers, and even trailers! In the city, these bikes are great: there's less pollution, no traffic jams, and they can get everywhere. However, it's not always easy, these bikes don't carry as much goods as trucks and have a limited delivery range.
But wait, that's not all! A major challenge remains. Existing infrastructures, such as depots (where parcels are stored for delivery), are poorly placed, and our valiant postal workers spend more time cycling from the depot to their delivery areas than to carry out their tours. What a loss of efficiency! Imagine all that time wasted when they could be out delivering parcels with a smile.
Let's change the organisation! It's time to review the location of depots. By setting up new depots closer to the delivery areas, our postal workers will spend less time travelling to their delivery areas and more time delivering parcels.
Modelization
The image of downtown Bordeaux shows the current depot represented by an orange circle.
Even if downtown Bordeaux seems small, and the addition of new depots may seem irrelevant, the idea here is to create a general model that can be adapted to any city. Since we don't have a huge budget, we're working with what we've got and the data we've got!
The aim is to minimize the total distances covered by the postal agents from their depot to their tour location.
We assume, of course, that a tour will be served by the nearest depot. Our budget allows us to build exactly K depots.Tours are represented by a point which is the barycenter of the GPS coordinates of the addresses served by an agent. In the example, 13 tours are carried out during the day.
The map above shows a solution where tours are represented by a dot, the i on the dots represent depots, and if a tour is the same color as a depot, it means that this tour is served by that depot.
All the tours are potential depot locations.
The data in this image correspond to real data and are stored in the file cyclo_real.dat.Note: there's a blue dot that looks closer to a red depot than to a blue depot, but this is of course wrong, because the distances calculated on this map are the distances in the Bordeaux bikeway graph, not as the crow flies.
© Louis Triouleyre-Roberjot,
Siao Leu Phouratsamay, Nadia Brauner
Juillet 2024