2022
Liu, Yang; Feng, Tao; Shi, Zhuangbin; He, Mingwei
Understanding the route choice behaviour of metro-bikeshare users Journal Article
In: Transportation Research Part A: Policy and Practice, vol. 166, pp. 460-475, 2022, ISSN: 0965-8564.
Abstract | Links | BibTeX | Tags: Metro-bikeshare integration, Multinomial logit model, Route choice, Smart card data
@article{LIU2022460,
title = {Understanding the route choice behaviour of metro-bikeshare users},
author = {Yang Liu and Tao Feng and Zhuangbin Shi and Mingwei He},
url = {https://www.sciencedirect.com/science/article/pii/S0965856422002890},
doi = {https://doi.org/10.1016/j.tra.2022.11.006},
issn = {0965-8564},
year = {2022},
date = {2022-01-01},
journal = {Transportation Research Part A: Policy and Practice},
volume = {166},
pages = {460-475},
abstract = {Understanding the determinants of the route choice behaviour on a multi-modal transit network of metro and shared bike is important to improve personalized multimodal travel services. This paper attempts to analyse the route choice behaviour of metro-bikeshare users considering passengers’ socio-economic attributes and perceived congestion which is approximated by load status. An abstract integrated metro-bikeshare network (IMBN) is built with virtual nodes by aggregating shared bike stations within the walkable distance and abstract routes by aggregating optional paths for each OD pair. Using the metro- and shared bike smart- card data from Nanjing, China, the route sets of metro-bikeshare users were extracted from the IMBN. A multinomial Logit model (MNL) was then applied to investigate the determinants of route choice behaviour for two types of users, namely “return-enter” and “exit-lease”, respectively. The results show that the models with the load status attributes have a better performance than the models without these attributes. We found the sensitivity of “exit-lease” users to the train crowding is significantly greater than that of the “return-enter” users. “Return-enter” users have a higher perception of out-of-vehicle travel time (OVTT) than that of in-vehicle travel time (IVT), while the “exit-lease” users have the opposite perception. Besides, the change rate of shared bike inventory, departure time and whether he or she is a regular user also have a significant impact on route choice behaviour. The findings can help policymakers and system operators to improve the services and the efficiency of the multimodal transportation system.},
keywords = {Metro-bikeshare integration, Multinomial logit model, Route choice, Smart card data},
pubstate = {published},
tppubtype = {article}
}
Understanding the determinants of the route choice behaviour on a multi-modal transit network of metro and shared bike is important to improve personalized multimodal travel services. This paper attempts to analyse the route choice behaviour of metro-bikeshare users considering passengers’ socio-economic attributes and perceived congestion which is approximated by load status. An abstract integrated metro-bikeshare network (IMBN) is built with virtual nodes by aggregating shared bike stations within the walkable distance and abstract routes by aggregating optional paths for each OD pair. Using the metro- and shared bike smart- card data from Nanjing, China, the route sets of metro-bikeshare users were extracted from the IMBN. A multinomial Logit model (MNL) was then applied to investigate the determinants of route choice behaviour for two types of users, namely “return-enter” and “exit-lease”, respectively. The results show that the models with the load status attributes have a better performance than the models without these attributes. We found the sensitivity of “exit-lease” users to the train crowding is significantly greater than that of the “return-enter” users. “Return-enter” users have a higher perception of out-of-vehicle travel time (OVTT) than that of in-vehicle travel time (IVT), while the “exit-lease” users have the opposite perception. Besides, the change rate of shared bike inventory, departure time and whether he or she is a regular user also have a significant impact on route choice behaviour. The findings can help policymakers and system operators to improve the services and the efficiency of the multimodal transportation system.
Liu, Yang; Ji, Yanjie; Feng, Tao; Shi, Zhuangbin
A route analysis of metro-bikeshare users using smart card data Journal Article
In: Travel Behaviour and Society, vol. 26, pp. 108-120, 2022, ISSN: 2214-367X.
Abstract | Links | BibTeX | Tags: Metro-bikeshare integration, Smart card data, Spatiotemporal patterns, Travel route
@article{LIU2022108,
title = {A route analysis of metro-bikeshare users using smart card data},
author = {Yang Liu and Yanjie Ji and Tao Feng and Zhuangbin Shi},
url = {https://www.sciencedirect.com/science/article/pii/S2214367X21000880},
doi = {https://doi.org/10.1016/j.tbs.2021.09.006},
issn = {2214-367X},
year = {2022},
date = {2022-01-01},
journal = {Travel Behaviour and Society},
volume = {26},
pages = {108-120},
abstract = {Few studies have analyzed individuals’ travel route characteristics in the integrated metro and bikeshare network. Taking Nanjing, China, as a case study, this paper analyzes the combined travel route of metro-bikeshare users based on three-week metro- and shared bike smart card data. The smart card data provides both boarding/borrowing and alighting/returning location and time, which makes it possible to trace each combined mode user’s actual metro routes. By assuming that the bike routes are the shortest paths between a specific OD pair from smart card records, we extract the combined metro-bikeshare travel routes of each user by using a metro route reconstruction method. The results show that over 60% of the metro-bikeshare users, most of whom are youth adults (18–35 years old), only occasionally use this mode. Further, the spatial and temporal characteristics of the combined metro-bikeshare travel routes are analyzed visually across different age groups. We find that most of the combined metro-bikeshare routes concentrated on Metro Line 2 with different cycling access/egress routes; however, the routes mostly travelled by different age groups are significantly different. Young adults are mainly distributed in suburban and exurban areas. Adults are more likely to be metro-bikeshare commuters, with travel routes usually passing through the central area of the city. Besides, the routes mostly travelled by the elderly are longer than those by other groups, and highly coincide with the routes mostly travelled by adults during the rush hours. Finally, relevant policy implementations are proposed in conjunction with metro-bikeshare users’ travel route characteristics.},
keywords = {Metro-bikeshare integration, Smart card data, Spatiotemporal patterns, Travel route},
pubstate = {published},
tppubtype = {article}
}
Few studies have analyzed individuals’ travel route characteristics in the integrated metro and bikeshare network. Taking Nanjing, China, as a case study, this paper analyzes the combined travel route of metro-bikeshare users based on three-week metro- and shared bike smart card data. The smart card data provides both boarding/borrowing and alighting/returning location and time, which makes it possible to trace each combined mode user’s actual metro routes. By assuming that the bike routes are the shortest paths between a specific OD pair from smart card records, we extract the combined metro-bikeshare travel routes of each user by using a metro route reconstruction method. The results show that over 60% of the metro-bikeshare users, most of whom are youth adults (18–35 years old), only occasionally use this mode. Further, the spatial and temporal characteristics of the combined metro-bikeshare travel routes are analyzed visually across different age groups. We find that most of the combined metro-bikeshare routes concentrated on Metro Line 2 with different cycling access/egress routes; however, the routes mostly travelled by different age groups are significantly different. Young adults are mainly distributed in suburban and exurban areas. Adults are more likely to be metro-bikeshare commuters, with travel routes usually passing through the central area of the city. Besides, the routes mostly travelled by the elderly are longer than those by other groups, and highly coincide with the routes mostly travelled by adults during the rush hours. Finally, relevant policy implementations are proposed in conjunction with metro-bikeshare users’ travel route characteristics.
2020
Gan, Z.; Yang, M.; Feng, T.; Timmermans, H.
Understanding urban mobility patterns from a spatiotemporal perspective: daily ridership profiles of metro stations Journal Article
In: Transportation, vol. 47, iss. 1, 2020, ISSN: 15729435.
Abstract | Links | BibTeX | Tags: LCLU, Ridership patterns, Smart card data, Station clustering, Urban mobility
@article{Gan2020b,
title = {Understanding urban mobility patterns from a spatiotemporal perspective: daily ridership profiles of metro stations},
author = {Z. Gan and M. Yang and T. Feng and H. Timmermans},
doi = {10.1007/s11116-018-9885-4},
issn = {15729435},
year = {2020},
date = {2020-01-01},
journal = {Transportation},
volume = {47},
issue = {1},
abstract = {Smart card data derived from automatic fare collection (AFC) systems of public transit enable us to study resident movement from a macro perspective. The rhythms of traffic generated by different land uses differ, reflecting differences in human activity patterns. Thus, an understanding of daily ridership and mobility patterns requires an understanding of the relationship between daily ridership patterns and characteristics of stations and their direct environment. Unfortunately, few studies have investigated this relationship. This study aims to propose a framework of identifying urban mobility patterns and urban dynamics from a spatiotemporal perspective and pointing out the linkages between mobility and land cover/land use (LCLU). Relying on 1 month’s transactions data from the AFC system of Nanjing metro, the 110 metro stations are classified into 7 clusters named as employment-oriented stations, residential-oriented stations, spatial mismatched stations, etc., each characterized by a distinct ridership pattern (combining boarding and alighting). A comparison of the peak hourly ridership of the seven clusters is conducted to verify whether the clustering results are reasonable or not. Finally, a multinomial logit model is used to estimate the relationship between characteristics of the local environment and cluster membership. Results show that the classification based on ridership patterns leads to meaningful interpretable clusters and that significant associations exist between local LCLU characteristics, distance to the city center and cluster membership. The analytical framework and findings may be beneficial for improving service efficiency of public transportation and urban planning.},
keywords = {LCLU, Ridership patterns, Smart card data, Station clustering, Urban mobility},
pubstate = {published},
tppubtype = {article}
}
Smart card data derived from automatic fare collection (AFC) systems of public transit enable us to study resident movement from a macro perspective. The rhythms of traffic generated by different land uses differ, reflecting differences in human activity patterns. Thus, an understanding of daily ridership and mobility patterns requires an understanding of the relationship between daily ridership patterns and characteristics of stations and their direct environment. Unfortunately, few studies have investigated this relationship. This study aims to propose a framework of identifying urban mobility patterns and urban dynamics from a spatiotemporal perspective and pointing out the linkages between mobility and land cover/land use (LCLU). Relying on 1 month’s transactions data from the AFC system of Nanjing metro, the 110 metro stations are classified into 7 clusters named as employment-oriented stations, residential-oriented stations, spatial mismatched stations, etc., each characterized by a distinct ridership pattern (combining boarding and alighting). A comparison of the peak hourly ridership of the seven clusters is conducted to verify whether the clustering results are reasonable or not. Finally, a multinomial logit model is used to estimate the relationship between characteristics of the local environment and cluster membership. Results show that the classification based on ridership patterns leads to meaningful interpretable clusters and that significant associations exist between local LCLU characteristics, distance to the city center and cluster membership. The analytical framework and findings may be beneficial for improving service efficiency of public transportation and urban planning.
Liu, Y.; Ji, Y.; Feng, T.; Timmermans, H.
Understanding the determinants of young commuters’ metro-bikeshare usage frequency using big data Journal Article
In: Travel Behaviour and Society, vol. 21, 2020, ISSN: 2214367X.
Abstract | Links | BibTeX | Tags: Metro-bikeshare integration, Negative binomial regression, Smart card data, Transfer frequency, Young commuter
@article{Liu2020,
title = {Understanding the determinants of young commuters’ metro-bikeshare usage frequency using big data},
author = {Y. Liu and Y. Ji and T. Feng and H. Timmermans},
doi = {10.1016/j.tbs.2020.06.007},
issn = {2214367X},
year = {2020},
date = {2020-01-01},
journal = {Travel Behaviour and Society},
volume = {21},
abstract = {This paper examines the determinants of young commuters’ frequency of using public bikes as a feeder mode to/from metro. Using three-week metro- and public bike- smart card data from Nanjing, 1,154 metro-bikeshare commuters aged 18–35 were extracted. As possible factors influencing the use of the combined mode, individual and household socio-demographics, travel-related attributes and built environment characteristics were extracted from multi-source data. A negative binomial regression model was used to examine the effects of these factors on usage frequency. We found that young commuters are the biggest group using metro-bikeshare system. They use public bikes frequently to transfer to/from metro when the cycling time is less than 10 min and the transfer happens during the morning peak. Built environment characteristics also influence usage frequencies, with high-density bike facilities being related to higher cycling rates in inner areas, and residential /employment locations related to lower rates of cycling in the core areas. This suggests that different measures and policies designed to encourage the integrated use of metro-bikeshare should be put forward for different regions.},
keywords = {Metro-bikeshare integration, Negative binomial regression, Smart card data, Transfer frequency, Young commuter},
pubstate = {published},
tppubtype = {article}
}
This paper examines the determinants of young commuters’ frequency of using public bikes as a feeder mode to/from metro. Using three-week metro- and public bike- smart card data from Nanjing, 1,154 metro-bikeshare commuters aged 18–35 were extracted. As possible factors influencing the use of the combined mode, individual and household socio-demographics, travel-related attributes and built environment characteristics were extracted from multi-source data. A negative binomial regression model was used to examine the effects of these factors on usage frequency. We found that young commuters are the biggest group using metro-bikeshare system. They use public bikes frequently to transfer to/from metro when the cycling time is less than 10 min and the transfer happens during the morning peak. Built environment characteristics also influence usage frequencies, with high-density bike facilities being related to higher cycling rates in inner areas, and residential /employment locations related to lower rates of cycling in the core areas. This suggests that different measures and policies designed to encourage the integrated use of metro-bikeshare should be put forward for different regions.
2019
Gan, Z.; Feng, T.; Wu, Y.; Yang, M.; Timmermans, H.
Station-based average travel distance and its relationship with urban form and land use: An analysis of smart card data in Nanjing City, China Journal Article
In: Transport Policy, vol. 79, 2019, ISSN: 1879310X.
Abstract | Links | BibTeX | Tags: Average travel distance, Land use mix, Metro, Quantile regression, Smart card data, Spatial scale
@article{Gan2019b,
title = {Station-based average travel distance and its relationship with urban form and land use: An analysis of smart card data in Nanjing City, China},
author = {Z. Gan and T. Feng and Y. Wu and M. Yang and H. Timmermans},
doi = {10.1016/j.tranpol.2019.05.003},
issn = {1879310X},
year = {2019},
date = {2019-01-01},
journal = {Transport Policy},
volume = {79},
abstract = {Few studies based on large sample data have examined mobility patterns from a travel distance perspective and investigated the potential influence of urban form and land use on people's daily travel distances. This paper provides additional empirical insights into spatiotemporal urban mobility patterns and their relationship with urban form and land use using station-based average travel distances (ATDs). Drawing on smart card data of the Nanjing metro system, land use data and open source points-of-interest (POIs)data, we apply exploratory spatial data and quantile regression analysis to examine distance patterns and explore the potential effects of urban form and land use calculated at different spatial scales (i.e. 800 m, 2 km and 5 km)on the ATDs. By comparing mobility patterns between weekdays and weekends and for different times of day, our findings highlight that ATDs are not uniformly nor randomly distributed in space. Positive spatial autocorrelation is found for different time segments. The results of OLS and quantile regression models show a positive and robust relationship between ATDs and distances to the city center (DCs). The models also prove that land use mix (especially measured at the 2 km and 5 km scale)significantly affects ATDs, supporting the importance of land use mix in decreasing daily travel distances. No significant relationship is found between ATDs and distances to the nearest subsidiary center (DSCs), while the employment/entertainment-residence balance has a marginal effect on ATDs at relatively large spatial scales (i.e. 2 km, 5 km). Consequently, with respect to reducing the ATDs, we recommend enhancing land use mix and reducing the imbalance between employment/entertainment and residence at larger spatial scales. Potential applications and future research directions are discussed. The findings in the present paper are helpful for guiding urban planning and policy making.},
keywords = {Average travel distance, Land use mix, Metro, Quantile regression, Smart card data, Spatial scale},
pubstate = {published},
tppubtype = {article}
}
Few studies based on large sample data have examined mobility patterns from a travel distance perspective and investigated the potential influence of urban form and land use on people’s daily travel distances. This paper provides additional empirical insights into spatiotemporal urban mobility patterns and their relationship with urban form and land use using station-based average travel distances (ATDs). Drawing on smart card data of the Nanjing metro system, land use data and open source points-of-interest (POIs)data, we apply exploratory spatial data and quantile regression analysis to examine distance patterns and explore the potential effects of urban form and land use calculated at different spatial scales (i.e. 800 m, 2 km and 5 km)on the ATDs. By comparing mobility patterns between weekdays and weekends and for different times of day, our findings highlight that ATDs are not uniformly nor randomly distributed in space. Positive spatial autocorrelation is found for different time segments. The results of OLS and quantile regression models show a positive and robust relationship between ATDs and distances to the city center (DCs). The models also prove that land use mix (especially measured at the 2 km and 5 km scale)significantly affects ATDs, supporting the importance of land use mix in decreasing daily travel distances. No significant relationship is found between ATDs and distances to the nearest subsidiary center (DSCs), while the employment/entertainment-residence balance has a marginal effect on ATDs at relatively large spatial scales (i.e. 2 km, 5 km). Consequently, with respect to reducing the ATDs, we recommend enhancing land use mix and reducing the imbalance between employment/entertainment and residence at larger spatial scales. Potential applications and future research directions are discussed. The findings in the present paper are helpful for guiding urban planning and policy making.