This session includes papers that investigated user behaviors in urban settings. The papers presented in this session focus on pedestrians, cyclists, and users of public transport and transportation network companies (TNCs). Research on intermodality, public transport mapping, equity, diffusion of the COVID-19, and road safety are presented.
Improving the Quality and Cost Effectiveness of Multimodal Travel Behavior Data Collection: A Case Study
Sean Barbeau, University of South FloridaShow Abstract
Cagri Cetin, University of South Florida
Multimodal transportation such as transit, bike, walk, ride-hailing (e.g., Uber, Lyft), carshare, and bikeshare are vital to supporting livable communities. However, current multimodal travel behavior data collection techniques, including travel behavior survey apps, have limitations (e.g., negative impact on battery life, user acquisition). This paper describes a case study of software developed to collect multimodal travel behavior data on an ongoing basis from users of an existing open-source mobile app for multimodal information, OneBusAway. To overcome battery life challenges, the research team used the Android Activity Transition API, which leverages hardware advancements in modern mobile phones. An update to the app was released to 676 users. Over 10 weeks, 74 users opted into the study without any incentive and contributed 65,582 trips. Key concerns for data collection when conserving battery life are the timeliness and accuracy of data. Location data was collected for 86% of all origins and destinations. Most delays in location acquisition when starting or ending activities were under a few minutes (e.g., 90th percentile of delay at origins was 3.2 minutes, 68th percentile was 14 seconds). The locations for origins and destinations were building-level accuracy or better (95th percentile of estimated accuracy was 48 meters). The primary cause of low activity classification confidence values seems to be uncertainty for walking vs. standing still. The software deployed in this project is a promising new tool with a tradeoff of reduced data density for the ability to collect data from many users for longitudinal studies with little incentives required.
Quantifying the impact of street lighting and walk paths on street inclusiveness: The case of Delhi
Laila AitBihiOuali (firstname.lastname@example.org), Imperial College LondonShow Abstract
Corentin Laffitte, Imperial College London
Daniel Graham, Imperial College London
This paper estimates the impact of street features on urban spaces inclusiveness near bus stops in Delhi, India. We mobilise a unique time-stamped and geo-coded panel dataset constructed on urban audits carried out by SafetiPin, a safety audit application for women which rates how safe streets in cities worldwide are over a large set of dimensions. A panel ordered probit model is used to assess the effect of street characteristics around bus stops on the number of people and the gender composition of the crowd for the period 2013-2019. We find that the level of crowding and the gender composition of bus stop surroundings (500m) are both significantly associated with the design urban spaces even when controlling for area unobserved heterogeneity and common time trends. Our results suggest that increasing the lighting from none to at least clear visibility increases the probability to have a crowded street (more than 10 people) by 8%. In addition, implementing a good quality walking path increases by 6% the probability to have more than 10 people and by 3% the probability to have a very crowded street. Regarding street inclusiveness, we find that increasing lighting increase the gender diversity in the street. More precisely, increasing lighting from ‘None’ to ‘Brightly Lit’ increases the probability to of the street to have some women and children by 1.5%, and increases by 2.9% the probability to have at least as many women as men in the street. This provides grounds for government intervention to boost the inclusiveness of streets through changes in the built environment.
Mapping the Intercounty Transmission Risk of COVID-19 In New York State Via Historical Commute Data
Shunhua Bai, University of Texas, AustinShow Abstract
Junfeng Jiao, University of Texas, Austin
Yefu Chen, University of Texas, Austin
The novel coronavirus disease of 2019(COVID-19) started at the end of 2019 has developed into a worldwide pandemic. As an epic center of the epidemic outbreak, New York state in the U.S. is facing life-and-death moments. The answer to how to distribute limited medical resources to the disastrous areas as well as to the risky areas is critical in local disease control practice. This paper investigates the COVID-19 transmission risk between different counties in the state of New York during the first month of the pandemic. This study constructs a comparable measure of COVID-19 outbreak status in different counties and quantifies their levels of severity. The study then conducts a transportation network analysis using historical inter-county commute data to establish transportation connections between counties and measures the county’s ability to spread/receive the disease. Results show that Queens, Kings, Westchester, Nassau, Bronx, Suffolk, and New York were all identified as the outbreak centers. Queens, Kings, Westchester, New York, and Monroe were major spreaders because of high volume, bi-directional commuting patterns. Saratoga and Oneida spread fewer outgoing cases, while Suffolk and Bronx were more affected by incoming cases. All other counties in the state are regarded as “community spreaders” with relatively low inter-county commutes, among which Rockland, Richmond, Essex, Orange were at mid-levels of severity in the outbreak. besides current outbreak centers as listed in results, specific attention should be given to Monroe, Saratoga, and Oneida. This study shows the role of transportation in facilitating disease control practice.
Spatio-temporal Integration between the Bicycle-sharing Service and the Metro Transit: A Case Study in Shanghai, China
Qing Yu, Tongji UniversityShow Abstract
Weifeng Li (email@example.com), Tongji University
Dongyuan Yang, Tongji University
Dockless bicycle-sharing is one of the novel transportation modes that emerged in recent years. As the newly-arisen service mode, it generates some new problems such as chaotic parking, low utilization rate, and so on, but it offers opportunities to solve the last-mile problem of public transit. To make the dockless bicycle-sharing service more effective and acceptable in the connection with public transit, it is necessary to implement different strategies of management according to the relationship between the user behavior of bicycle-sharing system and passenger flow public transit system. In this paper, the data of bicycle-sharing trips and the data of metro trips are utilized to assess the spatio-temporal integration between the two systems. A methodology is proposed to identify the cycling trips associated with metro stations, determine the cycling attraction area of metro stations and divide the policy zones for differentiated strategies of management. The proposed methodology is applied to a case study in Shanghai. According to the spatio-temporal integration between the passenger flow of metro stations and the bicycle-sharing use near metro stations, the metro stations in Shanghai are classified into four clusters, including stations with effective integration, stations with potential bicycle-sharing market, stations with deficiencies in bicycle-sharing use, and stations with rooms for improvement. Suggestions are made on the strategies of management in the different zones.
Do Public Transit and Agglomeration Economies Collectively Enhance Low-skilled Job Accessibility in Portland, OR?
Seunghoon Oh, University of CincinnatiShow Abstract
Na Chen, University of Cincinnati
The spatial mismatch hypothesis states that the decentralization of jobs undermines socially disadvantaged group’s job accessibility. One potential solution is the promotion of transit-induced agglomeration economies which could enhance job accessibility for low-wage or low-skilled workers by centralizing jobs to urban areas with relatively high-density surrounding transit stations. However, the gentrification through Transit-Oriented Development could limit jobs available to the underrepresented workers in the urban central area. In this case, agglomeration may not help these workers have better access to matching jobs. To unravel these relationships, this research investigates the effect of transit-induced agglomeration on gravity-based job accessibility by transit in Portland, OR. Furthermore, this study compares the agglomeration impact on low-skilled job accessibility to the overall job accessibility. Effective density is proxy for transit-induced agglomeration economies in the analysis. The relationship is estimated with spatial econometric models to evaluate the direct effect as well as the spillover effect of agglomeration. Overall, transit-induced agglomeration brings a positive effect on job accessibility. However, agglomeration economies of one location reduces the job accessibility of their neighbors, implying that concentrating economic activities beyond a certain spatial scale may absorb job opportunities from the neighboring areas. Standardized regression results suggest that the agglomeration effect on job accessibility is weaker for low-skilled workers than all workers, calling for decision-makers’ attention to the importance of “redistributing” the agglomeration effect through progressive land use and transportation policy efforts for underrepresented workers.
Collaborative Mapping of Urban Transport in Cartagena, Colombia
C. Erik Vergel-Tovar, Los Andes UniversityShow Abstract
Mónica Villegas Carrasquilla, Fundacion Corona
Maria Claudia Peñas, Former Director Cartagena Como Vamos
Daniel Toro Gonzalez, Universidad Tecnologica de Bolivar
Leonardo Canon Rubiano, The World Bank
Eliana Salas, Cartagena Como Vamos
Paulo Martinez, Universidad del Rosario
Mapping urban transport is an emerging methodology that enables researchers and citizens to map transit networks through participatory processes that enables to understand travel patterns, location of formal and informal transit networks and the geography of origin and destination points within urban areas. In this paper, we provide a detailed description of the methodology applied to map the urban formal and informal transit network of the city of Cartagena in Colombia. The visualization of the entire network at the urban agglomeration level provides information regarding relationships between urban growth and transit supply. The spatial data analysis also provides the opportunity to identify transit deserts, which correspond to informal settlements, increasing social exclusion issues for disadvantage groups. The methodology described in this paper also seeks to disseminate the procedures to scale-up the mapping methodology in other cities with high levels of informal transit. The generation of open data as part of this type of mapping methodologies aim to promote the data provision that engage citizens in the generation of new information to inform public policy. The use of information and communication technologies ICT can certainly support transportation planning processes by including data from informal transit and travel patterns to address social and economic issues associated to urban transport in cities with rapid urban growth.
Assessment of Non-Linear Associations between Area-Level Variables and Pedestrian Crash Frequency using a Model-based Gradient Boosting Framework
Dibakar Saha (firstname.lastname@example.org), Florida Atlantic UniversityShow Abstract
Eric Dumbaugh, Florida Atlantic University
This paper presents a study that evaluates the nature of the associations (i.e., linear or non-linear) between area-level variables and pedestrian crash frequency in urban areas. A number of variables representing demographic, socioeconomic, traffic, road network, and land use characteristics were considered to examine their associations with pedestrian crash frequency. The analysis was done based on three years (2015-2017) of pedestrian crashes in census block groups of Broward and Miami-Dade Counties in Florida. A machine learning approach, called the componentwise model-based gradient boosting algorithm that provides the flexibility to incorporate different base-learners, was implemented for the study analysis. In addition to a decision tree base-learner, the most common base-learner used in previous studies for evaluating safety, a generalized additive model (GAM) base-learner was used in this study to examine non-linear effect of predictor variables and a Markov random field base-learner was used to account for spatial correlation among analysis units. Models fitted with GAM base-learner were found to perform better than the models fitted with decision tree base-learners. Various types of linear and non-linear associations were found to exist between the variables and pedestrian crash frequency. Variables were also ranked in terms of their contributions in reducing bias in crash predictions. The study provides useful insights on how the results can help planners and policy makers to optimize the safety measures for pedestrians.
Results of the First Large-scale Survey of TNC Use in the Bay Area
Mark Bradley, RSG IncShow Abstract
Elizabeth Greene, RSG Inc
Bhargava Sana, San Francisco County Transportation Authority (SFCTA)
Drew Cooper, San Francisco County Transportation Authority (SFCTA)
Joe Castiglione, San Francisco County Transportation Authority (SFCTA)
Shimon Israel, Metropolitan Transportation Commission (MTC)
Christopher Coy, National Grid
Transportation Network Companies (TNCs) such as Uber and Lyft have grown tremendously over the past decade, particularly in the San Francisco Bay Area. Still, relatively little publicly available data exists about the users of these services, their travel behaviors, the volume of use, times and locations of TNC trips, and how TNC services are impacting transportation system performance overall. This paper describes the methods and descriptive results of the first large-scale smartphone-based TNC user survey conducted in the California Bay Area in fall 2018 and spring 2019. The study design took a standard household travel survey approach and optimized it to capture more TNC users and TNC travel behavior, using innovative methods to sample more TNC users via a targeted address-based sampling strategy and to capture more TNC trips by using a seven-day long travel period. The study also used innovative methods for the data weighting approach to ensure the resulting dataset was representative for the nine-county Bay Area across all days of the week. Lastly, the study methods were pre-tested and improved to facilitate a scalable project when applying the study methods to subsequent 2019 surveys in two other regions of California. The resulting datasets provide high quality data to support travel model estimation and calibration for trip and tour mode choice models, TNC user demographic profiles, and overall travel behavior analyses. The paper includes results describing mode shares, TNC trip purposes, TNC mode-choice substitution effects, time-of-day and day-of-week travel patterns across the nine-county Bay Area.
Impacts of School Reopening on Variations in Local Bus Performance in Sydney
Hudson Yao, University of SydneyShow Abstract
Wenbo Yan, University of Sydney
Linji Chen, University of Sydney
Emily Moylan, University of Sydney
Hema Rayaprolu, University of Sydney
During the COVID-19 pandemic, stay-at-home orders in conjunction with working from home, school closures and event cancellations resulted in decrease in travel demand. Under normal circumstances, these activities are components of trip chains and utilise a multi-modal transport network. The overall performance of the network can be traced through delays in the bus system as buses capture the changes in ridership and fluctuations in mixed traffic conditions. This paper explores the hypothesis that resumption of a single component in trip chains (i.e. school re-opening) is sufficient for a measurable change in transport system performance. This paper uses school reopening in Sydney, Australia as a case study to explore whether school-related trips affect bus system performance directly with higher student patronage or indirectly with heavier road congestion from parental car travels. Both stop dwell times and differences in delays between successive stops are used as measures of bus service indicators. Dwell times reflect the travel demand for buses and delay differences capture local changes in service reliability. We find that the increase in ridership has limited impacts on bus punctuality. Meanwhile, the level of local bus performance worsened after schools reopened, and the effect was more pronounced in commercial areas in the afternoon when schools end, suggesting secondary trip purposes such as leisure and shopping in addition to school pick-ups. This study reveals the interaction between different trip purposes during the post-COVID-19 period and throws light on the changes in travel behaviour patterns as travel restrictions are relaxed in pandemic situations.
Use of Exclusive and Pooled Ridehailing Services in Three Mexican Cities
Joanna Moody (email@example.com), Massachusetts Institute of Technology (MIT)Show Abstract
Enrique Esparza-Villarreal, Massachusetts Institute of Technology (MIT)
David Keith, Massachusetts Institute of Technology (MIT)
The global expansion of ridehailing platforms has been accompanied by a diversification of service offerings as platforms fit within new urban contexts. While ridehailing has been of great interest to transportation researchers, analysis of its adoption and use in cities in the Global South is lacking. To help address this knowledge gap, this study analyzes primary survey data collected from frequent users of DiDi Chuxing’s ridehailing platform in three Mexican cities: Mérida, Toluca, and Aguascalientes. We investigate how ridehailing fits into the travel behavior of its users, explicitly differentiating the express (exclusive) and comparte (pooled) service offerings. We find that (i) frequent use of ridehailing is positively correlated with use of public transport—city-run and privately-operated buses—and taxi, but negatively correlated with use of private car and motorcycle; and (ii) ridehailing trips are more likely to substitute public transport and taxi trips, but that the mode substitution depends on the service offering with high substitutability between express and comparte. This degree of substitutability suggests that there is potential to encourage ridehailing users to pool trips, increasing the occupancy rate of ridehailing vehicles and reducing their negative impacts on congestion. Among the many factors involved in choosing between exclusive and pooled services, our sample of ridehailing users in Mexico rate safety, travel time, travel time reliability, and price as key determinants, with a highly elastic relation between travel time and price. These results inform efforts by urban transportation policymakers and ridehailing operators to encourage pooling in the Latin American context.
Towards A Framework for Assessing the Fair Distribution of Space in Urban Streets
Gabriel Lefebvre-Ropars (firstname.lastname@example.org), Ecole Polytechnique de MontrealShow Abstract
Catherine Morency, Ecole Polytechnique de Montreal
Paula Negron-Poblete, Universite de Montreal
The increasing popularity of street redesigns highlights the intense competition for street space between their different users. More and more cities around the world mention in their planning documents their intention to rebalance streets in favor of active transportation, transit and vegetation. However, few efforts have managed to formalize quantifiable measurements of the balance between the different users and usages of the street. This paper proposes a method to assess the balance between the three fundamental dimensions of the street – the Link, the Place and the Environment – as well as a method to assess the adequation between supply and demand for the Link dimension at the corridor level. A series of open and government georeferenced datasets are integrated in order to determine the detailed allocation of street space for 11 boroughs of the city of Montréal, Canada. Travel survey data from the 2013 Origine-Destination survey is used to model different demand profiles on these streets. The balance between the three dimensions of the street is found to be most unbalanced in the central boroughs of the city, which are also the most dense and touristic neighborhoods. A discrepancy between supply and demand for transit users and cyclists is also observed across the study area. This highlights the potential of using a distributive justice framework to approach the question of the fair distribution of street space in an urban context.
DISCLAIMER: All information shared in the TRB Annual Meeting Online Program is subject to change without notice. Changes, if necessary, will be updated in the Online Program and this page is the final authority on schedule information.