Using AI to forecast waste collection times
A solution to predict waste collection time thanks to AI, and alert building caretakers via text messaging
// OUTCOMES & KEY RESULTS
A text message is sent to registered building caretakers before the collections truck arrive and another one once the truck has collected the bin. Thus buildings caretakers can know exactly when to take in an out the bins. The algorithm also improves its prediction over time with continuous data incoming from the microchips embed in the bins themselves.
The experimentation took place in Paris' 14th district and has been implemented since.
// PARTNERS INVOLVED
Partners provided the following data sets: digital collection routes, truck GPS coordinates, bin number, collection date and time, GPS location. The selected startup - Craft.ai - provides AI as a service and automated machine learning tools for individuals and companies.
// PAIN POINT & OPPORTUNITY
Almost 500 trucks pull out of waste processing facilities to collect rubbish on the streets of Paris every day, but bins cluttering public areas (pavements, entrance halls, etc.) are still common, either because the trucks have not yet emptied them or because nobody has returned them to storage.
This solution can be scaled to any city with a similar waste collection system. The solution can also be an opportunity for companies such as SUEZ and could become a new feature of offer to differenciate services.