Modeling on-board crowding contributions in public transportation systems using automated data sources
Time: Mon 2025-04-07 14.00
Location: U21, Brinellvägen 28A, Stockholm
Video link: https://kth-se.zoom.us/j/68852520133
Language: English
Subject area: Transport Science, Transport Systems
Doctoral student: Anastasios Skoufas , Transportplanering
Opponent: Docent Åse Jevinger, Malmö Universitet
Supervisor: Professor Erik Jenelius, Transportplanering, Centrum för transportstudier, CTS; Visiting Professor Oded Cats, Centrum för trafikforskning, CTR, Centrum för transportstudier, CTS, Transportplanering; Dr Matej Cebecauer, Transportplanering; Docent Wilco Burghout, Transportplanering
QC 20250318
Abstract
Cities worldwide are progressively attracting more residents, making the transportation supply provision challenging and the overcrowding phenomenon a new norm. Crowding negatively affects passengers’ travel experience and the operations of the public transportation system. So far, little attention has been given to how specific passenger groups, including the new residents of a city, contribute to public transportation crowding. Empirical knowledge of passenger groups’ impact on the crowding conditions in the system can guide tailored policy initiatives such as new fare structures, dedicated public transportation services, or infrastructure investments. Automated data sources in the public transportation sector can play an important role in this direction by i) offering opportunities for passenger segmentation and ii) providing data with high spatiotemporal resolution.
Paper I proposes a method based on smart card data for quantifying crowding contributions from a selected passenger group on the rest of the passengers at the journey level. We propose two novel metrics: time-weighted contribution to load factor and maximum contribution to load factor. The method is applied to two passenger groups: school students and passengers traversing Stockholm’s inner city. Results indicate that school students utilize 15% of the seating capacity in the Stockholm County case study area. Moreover, passengers traversing the inner-city occupy more than 80% of the seating capacity on the most affected network segment. The commuter rail corridor and its surrounding areas are found to be primarily affected by both selected passenger groups. Results can guide policy-making towards more efficient demand management and lower overall crowding conditions on the public transportation system.
Paper II extends the method proposed in Paper I in the context of new urban developments. The method captures the difference in crowding contributions induced by a newly developed area at the segment level. The method is applied to various urban developments, considering the classification of their types. Characteristics of the selected urban development categories such as the type, size, location, proximity to high-capacity public transportation connections, and socioeconomic characteristics are also concerned. Results reveal that the size, type of urban development, and proximity to a high-capacity connection are highly influential factors in determining the value and shaping the geographical extent of the crowding implications, regardless of its category. In addition, income and car ownership levels in the newly developed areas have a two-fold effect on shaping network-wide crowding contributions in terms of value and geographical spread. Results from Paper II can be incorporated into assessment frameworks for public transportation investments related to new urban developments. Last, results may assist in placing future urban developments accounting for the resulting crowding effects, therefore assisting towards more efficient public transportation networks and cities.