Master Thesis Defense - February 18, 2020: Decision Making for Urban Mobility: a Macro, Meso and Micro Analysis

Omar de Jesús Lucas Torres

Tuesday, February 18, 2020 at 10:00 a.m.
Room EV003.309

You are invited to attend the following M.A.Sc. (Quality Systems Engineering) thesis examination.

Examining Committee

Dr. J. Y. Yu, Chair
Dr. A. Awasthi, Supervisor
Dr. J. Yan, CIISE Examiner
Dr. C. Lai, External Examiner (ECE)

Abstract

Urban congestion is a challenge that cities commonly suffer across the globe. Traffic congestion and longer commutes are linked with poor cardiovascular and metabolic health, along with decreased energy and increased stress among the users. This is further translated in productivity and economic loss, increase in health service expenses and general decrease of quality of social wellbeing.

To improve this condition, the municipality administration has the role of implementing solutions to strategically address urban mobility. However, this is a complex task to achieve and normally involves limited resources, which make realworld deployments have a great inherited risk. Thus, decision-making is a task that has to be carefully addressed from different factors and scales. The thesis addresses the analysis of urban mobility in Montreal from three different levels of analysis.

At the macro level, the MTL Trajet dataset provides insight of mobility behaviour of participants through their trip coordinates. Using geometry datasets of quarter and boroughs of Montreal, the analysis is framed and processed via SQL and QGIS. Data visualization is presented in Choropleth maps, Flow maps and Chord diagrams using origin and destination of trips. Supporting processing task such as reverse geocoding to join attributes between datasets are used. Macroscopic analysis helps to identify a primary area of analysis seeking most transited region. The quarter of René-Lévesque in/and the borough of Ville-Marie are the most accessed areas in this study.

In the meso level, street network information from OpenStreetMap allows to make relations among the elements of the area, such as universities and their proximity of pedestrian zones. Resulting maps aid decision-making from a mesoscopic perspective, choosing the area of Concordia University as a suitable space for microscopic focus.

At the micro level, four areas of opportunity interpreted as transit policy-testing were identified. A custom microscopic network and synthetic demand for this area was used to simulate the impacts of these scenarios. The measures tested to improve urban mobility in the area are the restriction of street lanes for specific vehicle types and the inclusion of pedestrian areas. Experimentations with different levels of user modal share and shift are presented. This thesis approaches multiple tools for analytics on urban mobility using skills in SQL, R and Python, and open source software such as QGIS for spatial analysis and SUMO (Simulation of Urban Mobility) for microscopic simulation.

Results of macro, meso and micro analyses are included to provide recommendations for the administration of the city of Montreal. The inclusion of multiple restrained lanes for buses and high-occupancy vehicles around Concordia University and a pedestrian zone will allow to save time to road users, as long as single-passenger vehicle shifts towards public transit and shared–vehicles.