How to model the impact of logistics in order to ease traffic in cities?
La Poste handles 3/4 of parcel traffic and the number of parcels it carried grew by 3.6%
a year from 2009 to 2014, leading to a sharp rise in deliveries in cities. Requirements are also evolving (for example shoppers can now request delivery an hour after paying for their
These new demands are putting additional pressure on cities and entail the
need for specific arrangements to avoid the extra traffic worsening congestion.
At this point, there are several options to rearrange cities in order to ease traffic flows. But
these solutions rarely if ever accommodate in-city deliveries and their particular
The goal for this challenge is to provide a solution to harvest and escalate
delivery-related information that will furnish public-sector planners with the insights they
need to equip cities.
Use cases and experiment fields
The goal for this challenge is to develop a solution that:
- Classifies delivery vehicle inflows and outflows more accurately and by category (heavy vehicles, light vehicles, electric vehicles, etc.)
- Identifies the problems that delivery staff face (jammed roads, crowded or inconvenient delivery areas, etc.
Suggests improvements in difficult areas, for example:
– Resizing and relocating delivery bays
– Adjusting delivery timeframes (e.g. using outer lanes exclusively for deliveries from 2 to 5 pm)
– Multi-purposing streets, delivery areas and parking areas
Provides metrics to assess the impact of new measures such as the pollution-prevention plan, closing specific streets to traffic, restricting delivery timeframes, etc.
This experiment will take place in Paris, within an area to be defined in light of the tendered solution.
This experiment will take place in Paris, within an area to be defined in light of the tendered solution
- During the initial phase, La Poste and Paris City Council will supply:
– La Poste van delivery routes
– La Poste van geolocation information including:
- Document and parcel volumes by address
- Delivery categories (documents, parcels, etc.)
- Delivery addresses, doors with access touchpads, doors with electronic keys, buildings with caretakers
- Customer categories (residential or business)
- Important: this information is subject to strict confidentiality rules precluding disclosure of information such as specific drivers’ or recipients’ names
- A land registry map including delivery spaces
- During a possible second phase, a number of vehicles may be equipped with IoT sensors supplied by Sigfox to complete the existing data sets
- During a possible third phase, if initial results are conclusive, other partners could supply additional data to ensure the model is as representative as possible
There are no systems that effectively classify delivery traffic flows in cities today. This solution will open up opportunities for everyone involved in the programme:
- For La Poste and other logistics companies: an opportunity to factualise the constraints they are dealing with today and thereby substantiate the need for additional improvements that will ultimately shorten delivery times. This solution will also enable La Poste to support communities with expertise in in-city mobility and logistics
- For Setec and other engineering companies: an opportunity to complete its range of services for cities by including solutions that improve traffic conditions for delivery vehicles as well as private cars
- For the startup developing this solution: the goal is to develop a new product that Setec and La Poste can endorse, enhancing its credibility vis-à- vis local authorities
- For Paris City Council and other local authorities: an opportunity to reduce traffic and ease congestion resulting from logistics
This solution could also be used in other sectors (e.g. waste management with partner companies such as Suez).
La Poste provided Colisweb with 3 months of historical data including current delivery routes, types of customers, parcels, vehicles, etc. Colisweb then fed its algorithms, co-developed with INRIA French Insitute for Research in Computer Science and Automation , to create a solution that can be used to advise local authorities.
First, Colisweb was able to measure the impact of potenial urban modifications on logisics in terms of cost, time and CO2 per delivery. Second, the Colisweb solution was able to recommend the ideal leet size i.e. number of bikes, vans and trucks for a specific urban area.
Finally, the algorithms could be used to model the inter-combinaion of diferent delivery means like bikes and vans for last mile delivery.
Impact & Facts
First solution that accurately estimates the impact of city planning on logistics
– 32 vehicles and 77 postmen to cover the 1st and 2nd arrondissements
– 1 million parcels handled on average by La Poste each day
– 14 billion routes computed Go RAM and 6 CPUs used