Work zone Delay Modelling using Big Data and Machine Learning

Managing traffic through work zones is a highly complicated task where designers need to ensure smooth traffic flow without exposing workers to safety risks. One aspect of effective work zones traffic management involves predicting anticipated delays, understanding their causes, and proactively mitigating those causes. Unlike smooth traffic flow, predicting travel times and delays through interrupted traffic conditions is a more challenging task. This is particularly true when simultaneously managing traffic through multiple work zones, such as along the 700km Trans Mountain corridor between Edmonton and Vancouver where lapses in GPS signal and drops in cellular coverage are common. To help facilitate more effective work zone traffic management, this project aims to utilize Big Data collected along the Trans Mountain corridor to develop advanced statistical machine learning models capable of predicting travel times and delays along the corridor and other similar roads. The project will also investigate the impacts of several temporal, weather, and design variables on the delays to facilitate better traffic management.

Faculty Supervisor:

Suliman Gargoum

Student:

Partner:

ATS Traffic

Discipline:

Engineering

Sector:

Manufacturing

University:

The University of British Columbia - Okanagan

Program: