Smart Card Data Research Articles

Travel pattern recognition of urban rail passengers based on spatiotemporal sequence similarity

Understanding passenger travel patterns is helpful for the allocation of passenger resources in urban rail transit. Based on rail transit smart card data, this paper proposes a method to identify travel patterns by modeling individual spatiotemporal sequences. First, all the stations visited by individual passengers are extracted, and the similarity of the stations is calculated by the frequency of inter-station travel, the distance between stations and the activity duration of the stations. The main spatial activity area of the individual is divided using a hierarchical clustering algorithm. Then, the spatiotemporal sequence is inferred based on the individual's travel order. The sequence is a set of discrete values that characterize the spatiotemporal state. PCA-KL and K-Means++ are used to extract the similarity sequence structure to identify the passenger travel pattern. Finally, the rail transit smart card data of Xi'an in a certain month is taken as an example to identify its passenger travel pattern. The results show that complex passenger flow has 5 travel modes, of which 3 typical modes are commuting travel in a macro sense, accounting for 79% of the passenger flow. It can be seen that the pattern recognition based on the similarity of individual spatiotemporal sequences in this paper fully reflects the particularity and versatility of the research method, and is highly operational for different cities.

Managing transit-oriented development: A comparative analysis of expert groups and multi-criteria decision making methods

A key challenge for transport managers and planners in sustainable development is evaluating transit facilities' performance. While Multi-Criteria Decision-Making (MCDM) tools are often used, they can be influenced by experts' subjective biases. This study applies MCDM to assess Mass Rapid Transit (MRT) stations in Jakarta, Indonesia, focusing on Transit-Oriented Development (TOD). The primary goal is to compare stakeholder perspectives and MCDM methods, complemented by a sensitivity analysis and validation with real-world smart card data. The findings reveal significant differences in criteria weighting between Indonesian and non-Indonesian experts, and between academic and non-academic experts, especially in transit connectivity and land use diversity. The study also shows variations in station rankings across different MCDM methods. Sensitivity analysis identifies transit connectivity as the most critical criterion. Simple Additive Weighting (SAW) with linear normalisation aligns well with actual usage data and shows robustness in sensitivity analysis, making it the most reliable method for TOD evaluation. The study highlights the need for continuous TOD performance monitoring and the regular collection of real-world data on ridership and TOD indicators.

Evaluation of the Impacts of On-Demand Bus Services Using Traffic Simulation

This paper uses smart card data from Melbourne’s public transport network to model and evaluate the impacts of a flexible on-demand transport system. On-demand transport is an emerging mode of urban passenger transport that relies on meeting passenger demand for travel using dynamic and flexible scheduling using shared vehicles. Initially, a simulation model was developed to replicate existing fixed-schedule bus performance and was then extended to incorporate on-demand transport services within the same network. The simulation results were used to undertake a comparative analysis which included reliability, service quality, operational efficiency, network-wide effectiveness, and environmental impacts. The results showed that on-demand buses reduced average passenger trip time by 30%, increased vehicle occupancy rates from 8% to over 50%, and reduced emissions per passenger by over 70% on an average weekday compared to fixed-schedule buses. This study also offers insights for successful on-demand transport implementation, promoting urban sustainability. It also outlines future research directions, particularly the need for accurate short-term passenger demand prediction to improve service provision and passenger experience.

Activity-travel pattern inference based on multi-source big data

ABSTRACT We provide a comprehensive review of the literature on inferring activity-travel patterns (ATP) using multi-source big data; the increasing number of publications over time on this subject, demonstrates the importance of big data in this task. Our aims are to identify the advantages and research gaps in ATP inference and to promote further developments in this field. We clarify the fundamental concepts (i.e. ATP and its components), commonly used data sources, and inference processes employed in ATP inference studies. Emphasis is placed on two prominent big data sources: mobile phone data and smart card data. We outline the various approaches involved in the inference process, and we highlight existing shortcomings in data sources, ATP inference methodologies, and result validation. Based on the review, it is evident that future research should address several limitations in ATP inference. Firstly, it is necessary to improve the comprehensive understanding of ATP and understand the interrelationships among its different components. Secondly, it is necessary to integrate different data sources and leverage their respective strengths to gain deeper insights into activity-travel behaviour. Lastly, further investigation into emerging technologies such as artificial intelligence in ATP inference is warranted to improve inference accuracy. The findings of this study could provide valuable insights for policy makers, enabling them to gain a deeper understanding of activity-travel choice behaviour and develop more effective policies related to transportation system.

Unveiling mobility patterns beyond home/work activities: A topic modeling approach using transit smart card and land-use data

In this paper, a probabilistic topic modeling algorithm called Latent Dirichlet Allocation (LDA) is implemented to infer trip purposes from activity attributes revealed from smart card transit data in an unsupervised manner. While most literature focused on finding patterns for home and work activities, we further investigated non-home and non-work-related activities to detect patterns associated with them. Temporal attributes of activities are extracted from trip information recorded by Tehran subway’s automatic fare collection system. In addition, land-use data is also incorporated to further enhance spatial attributes for non-home/work activities. Various activity attributes such as start time, duration, and frequency in addition to land-use data are used to infer the activity purposes and patterns. We identified 14 different patterns related to non-commuting activities on the basis of both their temporal and spatial attributes including educational, recreational, commercial, and health and other service-related activity types. We investigated passengers’ activity pattern and behavior changes before and during COVID-19 pandemic by comparing the discovered patterns. For recreational patterns it is revealed that not only has the number of recreational patterns dropped, but also the duration of recreational activities decreased. Morning patterns of educational activities have also been eliminated and number of commercial activities was decreased during COVID-19. The proposed model demonstrates the ability to capture travel behavior changes for different disruptions using smart card transit data without performing costly and time consuming manual surveys which can be useful for authorties and decision makers.

Analyzing the reporting error of public transport trips in the Danish national travel survey using smart card data

AbstractHousehold travel surveys have been used for decades to collect individuals and households’ travel behavior. However, self-reported surveys are subject to recall bias, as respondents might struggle to recall and report their activities accurately. This study examines the time reporting error of public transit users in a nationwide household travel survey by matching, at the individual level, five consecutive years of data from two sources, namely the Danish national travel survey (TU) and the Danish smart card system (Rejsekort). Survey respondents are matched with travel cards from the Rejsekort data solely based on the respondents’ declared spatiotemporal travel behavior. Approximately, 70% of the respondents were successfully matched with Rejsekort travel cards. The findings reveal a median time reporting error of 11.34 min, with an Interquartile Range of 28.14 min. Furthermore, a statistical analysis was performed to explore the relationships between the survey respondents’ reporting error and their socio-economic and demographic characteristics. The results indicate that females and respondents with a fixed schedule are in general more accurate than males and respondents with a flexible schedule in reporting their times of travel. Moreover, trips reported during weekdays or via the internet displayed higher accuracies compared to trips reported during weekends and holidays or via telephone interviews. This disaggregated analysis provides valuable insights that could help in improving the design and analysis of travel surveys, as well accounting for reporting errors/biases in travel survey-based applications. Furthermore, it offers valuable insights underlying the psychology of travel recall by survey respondents.

Resilience of socio-technical transportation systems: A demand-driven community detection in human mobility structures

Existing scholarship on transportation resilience analysis has primarily focused on engineering resilience, often overlooking the intricate socio-technical dimensions. This oversight underscores the necessity for a more comprehensive understanding of the dynamic interplay between social, including travel behaviors, and technical infrastructure components within transportation systems. This article delves into the impact of “social shocks” on transportation systems, which are defined as disturbances affecting the social subsystem without yet affecting the technical subsystem. Drawing inspiration from C.S. Holling’s ecological resilience, which signifies a system’s ability to cope with change by adapting its structure and functionality, we propose a multi-level resilience assessment framework. It encompasses four mobility-related indicators: entropy (measuring network-level complexity), stationarity (assessing community compositional changes at the cluster level), and two node-level metrics — within-module degree and weighted participation coefficient — capturing location connectivity. These indicators proxy for evaluating the mobility structure and node functionality within the social subsystem. In a case study, we analyze historical smart card data to examine the mobility pattern’s structural changes within Hong Kong, a rail-oriented metropolis, during a prolonged and city-wide protest. The framework and associated indicators provide an alternative perspective for transit planners and operators, allowing them to assess both the overall system and individual stations, moving beyond traditional assessments of service supply and patronage changes.

An approach for two-dimensional convolutional neural networks for hourly passenger boarding demand prediction based on uneven smart-card data

An invaluable resource for understanding passenger boarding patterns and forecasting future travel demand is the tap-on smart-card data. Positive instances, on the other hand—boarding at a given bus stop at a certain time—are less common than negative instances when looking at the smart-card data (or instances) by boarding stops and by time of day. Machine learning algorithms that are used to estimate hourly boarding numbers at a certain location have been shown to be much less accurate when the data is imbalanced. Before using the smart-card data to forecast bus boarding demand, this research tackles the problem of data imbalance in the data. To create fake traveling instances to add into a synthetic training dataset containing more evenly distributed traveling and non-traveling examples, we suggest using deep generative adversarial networks (Deep-GAN). Next, a deep neural network, or DNN, is trained on the synthetic dataset to predict which instances from a given stop in a certain time frame will travel and which ones won't. According to the findings, resolving the data imbalance problem may greatly enhance the predictive model's functionality and make it more accurate in predicting ridership profiles. The suggested strategy may create a synthetic training set with a better similarity so diversity and, therefore, a stronger prediction capability, according to a comparison of the Deep-GAN's performance with other conventional resampling techniques. The study emphasizes the importance of the issue and offers helpful recommendations for enhancing the quality of the data and model performance for individual travel behavior analysis and travel behavior prediction.

Analyzing the transfer duration of public transport passengers using classification and regression tree-multiple-Cox proportional hazards (CART-Multi-Cox) model

ABSTRACT Transfer behavior is a critical factor influencing the travel efficiency of public transportation passengers. To address the potential group heterogeneity, the present work developed an integrated Classification and Regression Tree-Multiple-Cox Proportional Hazards (CART-Multi-Cox) model for transfer behavior analysis using smart card data in Shenzhen, China. Specifically, passengers are first grouped into different types based on transfer behavior features, and the influence of various independent variables on the transfer duration of different passenger groups is then examined. The results reveal that the proposed CART-Multi-Cox model is able to account for the heterogeneity effect and provides a deeper understanding about passengers’ transfer behavior and its underlying influencing mechanism. The findings offer valuable references for refined transfer behavior management and help enhancing the competitiveness of public transportation.

Is public transport a promising strategy for increasing physical activity? Evidence from a study of objectively measured public transport use and physical activity

BackgroundGreater public transport use has been linked to higher physical activity levels. However, neither the amount of physical activity associated with each daily public transport trip performed, nor the potential total physical activity gain associated with an increase in trips/day, has been determined. Using objective measures, we aimed to quantify the association between public transport use, physical activity and sedentary time.MethodsA longitudinal study of Australian adults living in Hobart, Tasmania, who were infrequent bus users (≥ 18 years; used bus ≤ 2 times/week). The number of bus trips performed each day was determined from objective smartcard data provided by the public transportation (bus) provider across a 36-week study timeframe. Accelerometer measured steps/day (primary outcome), moderate-to-vigorous physical activity (min/day), and sedentary time (min/day) were assessed across four separate one-week periods.ResultsAmong 73 participants across 1483 day-level observations, on days that public transport was used, participants achieved significantly more steps (β = 2147.48; 95%CI = 1465.94, 2829.03), moderate to vigorous physical activity (β = 22.79; 95% CI = 14.33, 31.26), and sedentary time (β = 37.00; 95% CI = 19.80, 54.21) compared to days where no public transport trips were made. The largest increase in steps per day associated with a one-trip increase was observed when the number of trips performed each day increased from zero to one (β = 1761.63; 95%CI = 821.38, 2701.87). The increase in the number of steps per day was smaller and non-significant when the number of trips performed increased from one to two (β = 596.93; 95%CI=-585.16, 1779.01), and two to three or more (β = 632.39; 95%CI=-1331.45, 2596.24) trips per day. Significant increases in sedentary time were observed when the number of trips performed increased from zero to one (β = 39.38; 95%CI = 14.38, 64.39) and one to two (β = 48.76; 95%CI = 25.39, 72.12); but not when bus trips increased from two to three or more (β=-27.81; 95%CI=-76.00, 20.37).ConclusionsGreater public transport use was associated with higher physical activity and sedentary behaviour. Bus use may yield cumulative increases in steps that amount to 15–30% of the daily recommended physical activity target. A policy and public health focus on intersectoral action to promote public transport may yield meaningful increases in physical activity and subsequent health benefits.

Intentional travel group and social network: Identification and dynamics during a pandemic

Using multiday continuous smartcard data in 2020, we investigate group-based travel in Hong Kong metro system by identifying metro riders intentionally traveling in groups (ITGs). ITGs serve as our proxies for citywide physical social interactions. Considering ITG members are interrelated through group-based trips, we construct a social network (an ITG network) formed by ITGs to explore the network properties and structures of ITG activities. Examining ITGs both before and during the COVID-19 pandemic, we measure the spatial patterns of ITGs and their dynamics across locales and over time. We find that the degree of the ITG network follows a heavy-tailed distribution. The network size and interconnections vary across time. Some ITG members are more influential vertices than others in maintaining the networks’ topological properties. We illustrate how new data and methods can be used to explore in-person interactions and social activity patterns in transit-reliant cities.

Unraveling the propagation dynamics of passenger congestion in public transportation systems using big smart card data

We propose a comprehensive framework to explore propagation of passenger crowdedness in public transportation systems using 74.16 million trip chains. A deep learning method HSTGCNT is adopted which firstly models normal spatio-temporal dependencies in time series data, and assigns an anomaly score to new data based on its prediction error. The derived discrete anomaly events are further aggregated into groups based on their spatio-temporal proximity by using a clustering algorithm that binds shared nearest neighbors together based on pairwise correlations. The clustering results uncover several key stages located in a propagation chain, such as a core event, adjacent events with spatio-temporal proximity, or a broadcasting group. Weibo data serves as a third-party dataset to verify the above anomaly clusters. Results show that the framework is effective to detect outlier clusters, with recall value being 0.86. Findings may assist public safety departments in taking pro-active actions to interrupt chaos from broadcasting.

Resilience-Based Approach to the Measurement of Headway Inconsistency

Maintaining a consistent headway is a primary goal in bus operation, but it poses a significant challenge because of fluctuating traffic conditions and uneven passenger demand. This leads to inconsistent headways and even bus bunching, resulting in negative consequences such as increased travel time and reduced vehicle capacity. Various control strategies have been developed to alleviate headway inconsistencies, but most of them have focused on whether an inconsistency occurs while overlooking its duration. This study proposes a resilience-based approach to measure headway inconsistencies considering both the depth and duration of the impacts with deviations from the optimal headway. The approach is devised to analyze the cumulative impacts of inconsistent headways comprehensively. It goes beyond addressing only extreme cases, such as bus bunching, and considers persistently uneven headways. We first identify vehicle trajectories from smart card data, then compute deviations from the center of the preceding and following vehicles as an index of impact. Using the concept of resilience, cumulative impacts are measured and their distribution is utilized to set the failure criterion. The approach is applied to 320 bus lines in Seoul to measure the deviations of headways and is used to analyze the characteristics of the bus lines and links where inconsistencies frequently occur. The proposed method enables transportation planners to identify factors that can lead to headway inconsistency, facilitating tailored and data-informed planning processes. This study provides a comprehensive framework that recognizes line-specific and link-specific factors, thereby contributing to enhancing the reliability and efficiency of public transportation systems.

Identifying Urban functional regions: A multi-dimensional framework approach integrating metro smart card data and car-hailing data

Urban functions often diverge from initial planning due to changes driven by residents’ behaviors. Effective urban planning and renewal require accurately identifying urban functional regions based on residents’ behavior data (including activity and travel data). However, previous methods have primarily relied on either point of interest (POI) data or a single source of traffic data, and often ignore the combined influence of residents’ activities and travel behaviors. In this study, we introduce a novel framework that integrates multiple sources of traffic data (such as metro smart card data and car-hailing data) with POI data to identify urban functional regions. This approach is unique because it simultaneously considers two critical dimensions of residents’ behavior: travel and activity behaviors. By combining these dimensions, we extract a comprehensive set of characteristics, including travel time, travel flow, origin-destination patterns, activity types, and activity time, which are then aggregated at the regional level (i.e., traffic analysis zone). To process these characteristics, we use latent Dirichlet allocation (LDA) to extract high-level semantic features from each data type. Additionally, to handle the sparse data from metro smart cards, we employ a specialized clustering technique. The integration of diverse and complementary information from multiple data sources enables more accurate and nuanced identification of urban functional regions than single data source and k-means clustering algorithm, providing valuable insights for urban planners.

Alighting location estimation from public transit data: a case study of Shenzhen

ABSTRACT This study proposes a framework to estimate alighting locations from Smart Card Data (SCD) that are absent information on entry-only public transport systems such as buses and trams. The proposed method uses the characteristics of SCD to (i) determine boarding locations from SCD and GPS-bus data based on exact match and time windows using common attributes, (ii) infer individuals’ home locations and user types from multimodal SCD, (iii) estimate alighting locations using inferred information with different scenarios such as with and without home locations based on the type of users. Reliable results are obtained once home locations are identified with high confidence for all user types. The proposed framework is applied to Shenzhen, China as a case study to validate the proposed model's effectiveness. The study offers valuable insight into aligning location estimation from user types to optimise the quality of public transport planning and services.

Evaluation and determinants of metro users' regularity: Insights from transit one-card data

Regularity is typically defined based on the repetitive travel behavior of individuals, referring to how often travelers would utilize a specific service within a given spatio-temporal context. However, previous research on metro users' regularity primarily utilized basic metric, for example metro trip frequency, to measure regularity. What's more, metro smart card data typically encompasses time, spatial features, and card type information, lacking individual attributes such as age, gender, and type of residence, which limits in-depth analysis correlating individual attributes with travel behavior. The study obtained transit one-card data from Nanjing, China, which enabled us to extract metro user's travel and individual information. Thus, the entropy rate methodology was employed to measure metro users' regularity, while machine learning techniques were used to analyze non-linear effects of built environment, travel-related, and individual attributes on regularity. Results indicate that the built environment, travel-related, and individual attributes account for 66.66%, 33.31%, and 0.03% of the total relative importance, respectively. Two most influential variables impacting regularity, namely entertainment POIs at the origin level (17.77%) and weekdays (17.51%), belong to the built environment and travel-related attributes, respectively. In terms of individual attributes, age exhibits a greater impact on regularity compared to gender and type of residence, manifested in the variation of regularity among different age groups. This finding can assist metro policymakers in understanding metro users' travel behavior, aiming to enhance operational efficiency and optimize the user experience.

Using smart card data to develop origin-destination matrix-based business analytics for bus rapid transit systems: case study of Jakarta, Indonesia

Bus rapid transit systems (BRT) have been an indispensable public transportation pillar, especially in densely populated regions. Accurate insight into the BRT network's utilization is vital in BRT resource allocation planning contexts. This research focuses on how operators can utilize passengers’ smart card data to develop origin-destination (OD) matrix-based business analytics. This research proposes a hybrid approach combining trip chaining, direct pairing, mode estimation methods, and visual analytics development. The novel approach is robust in handling incomplete smart card data transactions to generate origin-destination matrices and corresponding visual analytics as decision support systems for the BRT operators. As a case study, we applied and validated the proposed analytics to more than 20.6 million smart card transactions in one of the largest global BRT systems in Jakarta, Indonesia.

Comparative analysis of the optimal route recommendation models based on city public transportation network data

In recent years, studies on smart cities have attracted much attention in the literature. In the smart city concept, urban public transportation network studies are one of the most important issues. Afterthe 2000s, smart cards have been widely used in city public transportation systems. Using the data collected through the smart card, many models have been created in line with the smart city concept. In the literature, there are various optimal route recommendation models using urban public transportation network and city smart card data. In this study, some current models built on the basis of urban public transportation network and smart card data are discussed. The specifications of the models, their important differences, the topologies they use, optimization criteria, and computational complexities are analyzed. Dijkstra’s algorithm, which is widely used for the solution of optimal route recommendation models, and its various modifications are analysed. Additionally, various models developed by applying fuzzy logic are examined. Comparative analysis of public transportation network models is given.

Estimation of schedule preference and crowding perception in urban rail corridor commuting: An inverse optimization method

This paper introduces an inverse optimization method to uncover commuters’ schedule preference and crowding perception based on aggregated observations from smart card data for an urban rail corridor system. The assessment of time-of-use preferences typically involves the use of econometric models of discrete choice based on detailed travel survey data. However, discrete choice models often struggle with potential endogeneity issues in behavioral observations when estimating individual samples from massive transit data with limited exogenous identifying information. This motivates us to employ an equilibrium modeling approach to capture the dynamism hidden in commuters’ departure time decision-making from aggregations. Assuming user optimality in observed choices, an inverse optimization method is proposed to find a set of preference parameters in the stochastic user equilibrium-based morning commuting model with heterogeneous commuters so that the resulting equilibrium pattern best approximates the observed departure rate distribution over time. The proposed inverse optimization problem can be formulated by a bi-level programming model and a sensitivity analysis-based solution framework is further designed for model estimation. Lastly, the smart card data and train timetable data from the rail corridor along the Beijing Subway Batong Line are synthesized for a case study to estimate commuters’ departure time choice preferences during morning peak periods, as well as to validate the robustness and practicality of the proposed method.

Increasing the precision of public transit user activity location detection from smart card data analysis via spatial–temporal DBSCAN

Smart Card (SC) systems have been increasingly adopted by public transit (PT) agencies all over the world, which facilitates not only fare collection but also PT service analyses and evaluations. Spatial clustering is one of the most important methods to investigate this big data in terms of activity locations, travel patterns, user behaviours, etc. Besides spatio-temporal analysis of the clusters provide further precision for detection of PT traveller activity locations and durations. This study focuses on investigation and comparison of the effectiveness of two density-based clustering algorithms, DBSCAN, and ST-DBSCAN. The numeric results are obtained using SC data (public bus system) from the metropolitan city of Konya, Turkey, and clustering algorithms are applied to a sample of this smart card data, and activity clusters are detected for the users. The results of the study suggested that ST-DBSCAN constitutes more compact clusters in both time and space for transportation researchers who want to accurately detect passengers’ individual activity regions using SC data.