Search

Drawing on transit smart card data allied with local weather station records over a 12-month period, this paper takes Brisbane, Australia as the study context and examines the way in which weather impose influences on public transit passenger's travel behaviour. In terms of transit passenger's spatiotemporal origin–destination (OD) information, a new travel behaviour indicator called stickiness is developed to reflect passenger's travel similarity in using the transit service over a period. By applying a suite of regression models, it shows that weather's effects on transit passenger's travel behaviour vary by temporal period and passenger type. In general, weather is shown to exert a stronger effect on passenger's stickiness during midday off-peak hours in comparison with either AM peak or PM peak. Across all types of passengers, child passengers are found to be most tolerant to weather changes and even be stickier to their regular travel pattern during AM peak. The modelling results also reveal that when it relates to journey-to-work or journey-to-school, weather's effects on alighting location are much smaller than other three OD-related travel behaviour features (i.e., boarding time, boarding location, and alighting time). In contrast with the postponement during AM peak, transit passengers are prone to bring forward their trip back home during PM peak in response to poor weather conditions. Moreover, this paper verifies that weather parameters are not singly perceived by transit passengers but as an interrelated unity in shaping transit passenger's travel behaviour. By enriching the weather–travel behaviour scholarship, the empirical findings of this study are helpful to render a holistic understanding of the weather–travel behaviour relationship and have important implications for transit operators in building a more weather-resilient transit system. [ABSTRACT FROM AUTHOR]

This paper presents a data fusion method to infer the origin and destination zones of transit trips from smart card data. The fusion framework has disaggregate mixed multinomial logit models at its core that predict the most probable origin and destination zones of individual transit trips using smart card transaction records, land use data, and transit system characteristics. The logit models are estimated using revealed trip origin and destination responses from a travel survey that are augmented by land use and transit system data to provide contexts about the zones' trip generation and attraction potentials. For empirical analysis, the methodology is applied to the smart card system of the Greater Toronto and Hamilton Area. Specifically, it is tested under different system configurations (tap-on-only and tap-on-and-off) and for networks with substantial shares of automobile and walk access/egress. When applied to transit trips constructed from the smart card transactions, the estimated models successfully capture the spatial distribution of trip origin and destination at the traffic analysis zone level. The empirical analysis also demonstrates that the proposed fusion method can be appropriately used to reconcile information provided by transit smart card and travel surveys to generate up-to-date transit demand data necessary for public transport planning and operations. [ABSTRACT FROM AUTHOR]

Many human activities, including daily travel, show a mix of stable, intermittent and changing patterns in demand by individuals over time. However, the lack of continuous, long-term, passenger-linked data for public transport (PT) journeys means that we do not know how passenger ridership evolves in real-world networks. This paper proposes the CIAM model for classifying long-term passenger engagement with PT. CIAM is a data-driven model combining year-on-year churn (C), monthly intensity (I), annual (A) and multi-year (M) engagement. Parameter search algorithms are used to ensure that the learned features are distinctive and robust. We evaluated CIAM using a 5-year dataset from a PT network with over 300 million journeys. CIAM identified distinct patterns of long-term ridership at multiple time scales. Although the total number of annual journeys was relatively stable over the five years, we found long-term differences between passenger subgroups. Churn of passengers was a major factor in ridership with only 55% of passengers retained from year to year. Patterns of annual engagement are often intermittent, so short-term snapshots of a few weeks are typically not good indicators for longer term engagement. Only 27% of high-frequency, full-fare riders still have the same level of engagement four years later, compared with 55% who continue high-frequency engagement after only one year. • CIAM learns long-term patterns of PT engagement at multiple temporal scales. • Churn is a major factor of ridership with year on year passenger retention of only 55%. • Annual engagement is a mixture of contrasting short-term intensities. • Only 27% of high-frequency, full-fare passengers maintain that engagement for 5 years. • Data-driven CIAM identifies underlying context transitions from movement data alone. [ABSTRACT FROM AUTHOR]

Service bottlenecks are a key barrier to building a resilient public transport system. In this paper, we propose a new approach to automatically extract the role of a station in dynamical public transport flow networks based on the emerging role discovery method in network science. The term "role" in this study refers to the distinctive position or function that a station plays within the public transport flow network. Using smart card data from Nanjing public bikesharing agencies, we first construct dynamical public transport flow networks with notions of dynamical graph and edge. We then develop a dynamical algorithm to recursively compute the structural flow characteristics of nodes in passenger flow networks. Non-negative Matrix Factorization is conducted to extract the role memberships from the derived structural feature matrix and interpret each role in terms of measurements with practical values. The network hubs and potential service bottlenecks are then identified based on their operating characteristics and dynamics. Furthermore, the day-to-day and within-day role dynamics of public transport stations over time are unveiled. The results contribute to a better understanding of the interplay between stations in the network, and the identification of roles provides insight for public transport agencies to improve service resilience. [ABSTRACT FROM AUTHOR]

Travel activities influence individual travel location, time, and frequencies. Understanding individual travel regularity under different activities is crucial for individual mobility prediction and urban facilities planning. Existing studies need further improvement as they have not adequately taken into account the impact of activity factors on individual travel patterns. In this paper, we propose an innovative framework for exploring travel regularity and the correlation with urban region attributes at the individual level. The framework's novelty and uniqueness lie in its three layers: (i) A trip pre-processing layer for extracting every complete public transit trip and trip purposes (e.g. commuting activity and non-commuting activity) by combining individual's travel characteristics and public transport network. (ii) An entropy-based travel regularity measurement and K-means based multiple-view clustering layer to assess the extent of individual travel repetition over time for various travel activity and investigate the similarities and differences among users. (iii) A region dependence correlation analysis layer for exploring the correlation between individual travel regularity and regions attributes of two key locations: home and workplace More importantly, by employing the framework, we gained empirical insights based on a large-scale dataset (covering 0.64 million users) collected from public traffic smart cards. For instance, commuters can be categorized into four groups based on the regularity of their commuting activities: Regular Workers (24%), Flex-time Workers (21%), Overtime Workers (35%), and Other Workers (20%). The distribution of these four groups is associated with workplace's administrative characteristics, with a Pearson correlation of 0.503. In addition, individuals can also be classified into two groups based on their travel regularity of non-commuting activities: Limited Visitors (37%) and Active Explorers (63%). The distribution of these individuals is correlated with the coverage ratio of public transit facility in their home locations, showing a Pearson correlation of 0.604. • A framework exploring individual travel regularity under different purpose. • Entropy-Space K-means for travel activity-based individual classification. • Inter-personal and intra-personal variability of individual travel patterns. • Correlation between individual travel regularity and urban region attributes. [ABSTRACT FROM AUTHOR]

Most crowding measures in public transportation are usually aggregated at a service level. This type of aggregation does not help to analyze microscopic behavior such as exposure risk to viruses. To bridge such a gap, our paper proposes four novel crowding measures that might be well suited to proxy virus exposure risk at public transport. In addition, we conduct a case study in Santiago, Chile, using smart card data of the buses system to compute the proposed measures for three different and relevant periods of the COVID-19 pandemic: before, during, and after Santiago's lockdown. We find that the governmental policies diminished public transport crowding considerably for the lockdown phase. The average exposure time when social distancing is not possible passes from 6.39 min before lockdown to 0.03 min during the lockdown, while the average number of encountered persons passes from 43.33 to 5.89. We shed light on how the pandemic impacts differ across various population groups in society. Our findings suggest that poorer municipalities returned faster to crowding levels similar to those before the pandemic. • Four public transport crowding measures that proxy virus exposure risk are proposed. • A case study is conducted in Santiago, Chile, using GPS and smart card data. • Three periods, before, during, and after COVID-19 lockdowns are compared. • Governmental policies diminished crowding considerably during lockdowns. • Poorer municipalities returned faster to pre-pandemic crowding levels. [ABSTRACT FROM AUTHOR]

Smart card data (SCD) allow analyzing mobility at a fine level of detail, despite the remaining challenges such as identifying trip purpose. The use of the SCD may improve the understanding of transit users' travel patterns from precarious settlements areas, where the residents have historically limited access to opportunities and are usually underrepresented in surveys. In this paper, we explore smart card data mining to analyze the temporal and spatial patterns of the urban transit movements from residents of precarious settlements areas in São Paulo, Brazil, and compare the similarities and differences in travel behavior with middle/high-income-class residents. One of our concerns is to identify low-paid employment travel patterns from the low-income-class residents, that are also underrepresented in transportation planning modeling due to the lack of data. We employ the k-means clustering algorithm for the analysis, and the DBSCAN algorithm is used to infer passengers' residence locations. The results reveal that most of the low-income residents of precarious settlements begin their first trip before, between 5 and 7 AM, while the better-off group begins from 7 to 9 AM. At least two clusters formed by commuters from precarious settlement areas suggest an association of these residents with low-paid employment, with their activities placed in medium / high-income residential areas. So, the empirical evidence revealed in this paper highlights smart card data potential to unfold low-paid employment spatial and temporal patterns. [ABSTRACT FROM AUTHOR]

User profiles, considered as one of the fundamental inputs of recommendation systems and customized services, can be rationally applied in the public transport domain to represent passengers' characteristics and behavioral preferences. A user profile of a bus passenger, termed as a bus passenger profile (BPP), is an assortment of labels containing passengers' travel features. This paper proposes a data-driven framework for developing BPP so as to provide guidance on how to create and estimate user profiles for bus passengers based on smart card data. The proposed method comprises three steps. (i) Data preprocessing aimed at extracting key information and preparing passenger profiling. (ii) A tag system aimed at storing the estimated travel features of passengers. (iii) Knowledge graphs aimed at connecting various BPP with semantic edges for practical application of prior knowledge in downstream tasks. The developed framework is implemented in a case study of the Beijing bus system. Deployment of the developed framework has demonstrated that it can satisfactorily develop BPP, while prior knowledge from the BPP-based knowledge graphs can benefit downstream tasks. [ABSTRACT FROM AUTHOR]

In recent decades, transit-oriented development (TOD) has been considered as an effective way to alleviate the negative impacts of rapid urbanization. The ridership of rail stations is an important indicator for understanding the relationship between rail transit and land use, which helps enhance their coordinated development. How to regulate such development is a fundamental question that planning of rail stations and surrounding land use endeavors to answer. Understanding the relationship between rail transit and land use is the basis of effective planning. From smart card data (SCD), the boarding and alighting ridership time series can be extracted to understand their relationship from a new perspective. Current studies mainly consider the temporal similarity of ridership time series between individual stations but ignore spatial dependency between a station and its adjacent stations. To fill this gap, this paper proposes a novel method for refining the understanding from a spatiotemporal similarity perspective. First, we measure the temporal similarity of ridership time series between two individual stations using the autocorrelation function method. Second, we determine the spatiotemporal similarity between two stations by integrating the temporal similarity from their adjacent stations based on spatial dependency. Finally, the extracted spatiotemporal similarity is used for spectral clustering. As a case study, we analyze SCD on five consecutive weekdays at 289 rail stations in Shanghai, China. The land use functions of rail station areas are classified into 6 clusters: employment-oriented stations, residential-oriented stations, mixed stations, etc. By comparing the enrichment factors of each type of points of interest (POIs) in these clusters, the dominant land use function between our classification and POI information is demonstrated to be consistent. Furthermore, using three clustering performance indexes and several typical stations, the priority of our proposed method with integration of the spatial dependency can be confirmed to be superior to those without considering such important information. We conclude that the proposed method and findings are beneficial to balancing the needs of public transportation development and promoting the integration of transportation and land use for TOD implementation. • A novel method is proposed for inferring land use functions of station areas. • The proposed method is based on spatiotemporal similarity in ridership data. • Validation of the proposed method is verified by a case study in Shanghai. • The inferred land use of station areas spatially varies in Shanghai. • This study refines understanding of relationship between rail transit and land use. [ABSTRACT FROM AUTHOR]

In urban areas, public transport can improve sustainable transport by reducing vehicles and congestion, and improving accessibility. Mass rapid transit (MRT) is especially important for large cities, such as Taipei City, Taiwan. In particular, MRT aims to improve mobility and reliability, but has limitations in providing first and last mile accessibility to final destinations or origins. Taipei City, Taiwan has introduced a public bike sharing scheme to service this gap. However, few studies have addressed how a public bike sharing scheme assists first and last mile accessibility. This study uses Taipei City as a case study to investigate this issue by comparing the demand and supply of the public bike sharing scheme, YouBike, at a detailed spatial scale. The supply of YouBike is represented by the time saving compared to walking for each identified origin and destination pair in the study area (i.e., Point of Interests (POIs) and MRT stations). The demand for YouBike is total trips from each YouBike station to a MRT station using public transport smart card data. By comparing the demand and supply for over 400 zones or villages, service gaps and areas of unbalanced service can be identified. The results show that YouBike does provide first and last mile service for the MRT network with some evidence of service mismatch in the study area, i.e., high service for low demand and vice versa. The conclusions of the paper can help cities wanting to introduce a bike sharing schemes to improve first and last mile transport. [ABSTRACT FROM AUTHOR]

Predictions on Public Transport (PT) ridership are beneficial as they allow for sufficient and cost-efficient deployment of vehicles. On an operational level, this relates to short-term predictions with lead times of less than an hour. Where conventional data sources on ridership, such as Automatic Fare Collection (AFC) data, may have longer lag times and contain no travel intentions, in contrast, trip planner data are often available in (near) real-time and are used before traveling. In this paper, we investigate how such data from a trip planner app can be utilized for short-term bus ridership predictions. This is combined with AFC data (in this case smart card data) to construct a ground truth on actual ridership. Using informative variables from the trip planner dataset through correlation analysis, we develop 3 supervised Machine Learning (ML) models, including k-nearest neighbors, random forest, and gradient boosting. The best-performing model relies on random forest regression with trip planner requests. Compared with the baseline model that depends on the weekly trend, it reduces the mean absolute error by approximately half. Moreover, using the same model with and without trip planner data, we prove the usefulness of trip planner data by an improved mean absolute error of 8.9% and 21.7% and an increased coefficient of determination from a 5-fold cross-validation of 7.8% and 18.5% for two case study lines, respectively. Lastly, we show that this model performance is maintained even for the trip planner requests with prediction lead times up to 30 min ahead, and for different periods of the day. We expect our methodology to be useful for PT operators to elevate their daily operations and level of service as well as for trip planner companies to facilitate passenger replanning, in particular during peak hours. • Empirical analysis of correlation in bus ridership with short-term trip planner data (advance travel advice requests). • Several interpretable machine-learning prediction models with and without trip planner data are estimated. • Two case studies with different line characteristics show trip planner data can reduce mean absolute error by 9 to 22%. • Performance of best prediction model is robust across all periods of day and lead times up to 30 min ahead. • Developed prediction methodology is applicable to both public transport operators and trip planner providers. [ABSTRACT FROM AUTHOR]