Stop-level probabilistic demand forecasting for public transit services

This project aims to develop advanced statistical methods to forecast how many passengers will be waiting at bus stops at any given time, considering various factors such as weather and special events. This project will do two main things. First, we will forecast the overall number of passengers over a given period, such as an hour. This level of aggregate demand forecasting can help transit agencies to design bus routes and schedules. Second, we will forecast the passenger flow for each bus at every stop. This granular demand forecasting allows for real-time adjustments to manage crowding and improve the overall passenger experience. For our partner organization, this project offers the dual benefits of enhancing the passenger experience—making public transit a more attractive option for commuters—and contributing to environmental goals by optimizing operations.

Faculty Supervisor:

Lijun Sun;Martin Trépanier

Student:

Partner:

Exo

Discipline:

Engineering

Sector:

Transportation and warehousing

University:

McGill University

Program: