The station-free sharing bike demand forecasting with a deep learning approach and large-scale datasets

The station-free sharing bike is a new sharing traffic mode that has been deployed in a large scale in China in the early 2017. Without docking stations, this system allows the sharing bike to be parked in any proper places. This study aimed to develop a dynamic demand forecasting model for station-free bike sharing using the deep learning approach. The spatial and temporal analyses were first conducted to investigate the mobility pattern of the station-free bike sharing. The result indicates the imbalanced spatial and temporal demand of bike sharing trips. The long short-term memory neural networks (LSTM NNs) were then developed to predict the bike sharing trip production and attraction at TAZ for different time intervals, including the 10-min, 15-min, 20-min and 30-min intervals. The validation results suggested that the developed LSTM NNs have reasonable good prediction accuracy in trip productions and attractions for different time intervals. The statistical models and recently developed machine learning methods were also developed to benchmark the LSTM NN. The comparison results suggested that the LSTM NNs provide better prediction accuracy than both conventional statistical models and advanced machine learning methods for different time intervals. The developed LSTM NNs can be used to predict the gap between the inflow and outflow of the sharing bike trips at a TAZ, which provide useful information for rebalancing the sharing bike in the system.     

Factors influencing bike share membership: An analysis of Melbourne and Brisbane

The number of bike share programs has increased rapidly in recent years and there are currently over 700 programs in operation globally. Australia’s two bike share programs have been in operation since 2010 and have significantly lower usage rates compared to Europe, North America and China. This study sets out to understand and quantify the factors influencing bike share membership in Australia’s two bike share programs located in Melbourne and Brisbane. An online survey was administered to members of both programs as well as a group with no known association with bike share. A logistic regression model revealed several significant predictors of membership including reactions to mandatory helmet legislation, riding activity over the previous month, and the degree to which convenience motivated private bike riding. In addition, respondents aged 18–34 and having docking station within 250 m of their workplace were found to be statistically significant predictors of bike share membership. Finally, those with relatively high incomes increased the odds of membership. These results provide insight as to the relative influence of various factors impacting on bike share membership in Australia. The findings may assist bike share operators to maximize membership potential and help achieve the primary goal of bike share – to increase the sustainability of the transport system.     

Synopsis of bicycle demand in the City of Toronto: Investigating the effects of perception,consciousness and comfortability on the purpose of biking and bike ownership

This paper provides an empirical basis for the evaluation of policies and programs that can increase the usage of bikes for different purposes as well as bike ownership. It uses an integrated econometric model of latent variable connecting multiple discrete choices. Empirical models are estimated by using a bicycle demand survey conducted in the City of Toronto in 2009. Empirical investigations reveal that latent perceptions of ‘bikeability’ and ‘safety consciousness’ directly influence the choice of biking. It is also found that the choice of the level of bike ownership (number of bikes) is directly influenced by latent ‘comfortability of biking’. The number of bikes owned moreover has a strong influence on the choices of biking for different purposes. It is clear that bike users in the City of Toronto are highly safety conscious. Increasing on-street and separate bike lanes proved to have the maximum effects on attracting more people to biking by increasing the perception of bikeability in the city, comfortability of biking in the city and increasing bike users’ sense of safety. In terms of individuals’ characteristics, older males are found to be the most conformable and younger females are the least comfortable group of cyclists in Toronto.     

Free-floating bike sharing: Solving real-life large-scale static rebalancing problems

Free-floating bike sharing (FFBS) is an innovative bike sharing model. FFBS saves on start-up cost, in comparison to station-based bike sharing (SBBS), by avoiding construction of expensive docking stations and kiosk machines. FFBS prevents bike theft and offers significant opportunities for smart management by tracking bikes in real-time with built-in GPS. However, like SBBS, the success of FFBS depends on the efficiency of its rebalancing operations to serve the maximal demand as possible.Bicycle rebalancing refers to the reestablishment of the number of bikes at sites to desired quantities by using a fleet of vehicles transporting the bicycles. Static rebalancing for SBBS is a challenging combinatorial optimization problem. FFBS takes it a step further, with an increase in the scale of the problem. This article is the first effort in a series of studies of FFBS planning and management, tackling static rebalancing with single and multiple vehicles. We present a Novel Mixed Integer Linear Program for solving the Static Complete Rebalancing Problem. The proposed formulation, can not only handle single as well as multiple vehicles, but also allows for multiple visits to a node by the same vehicle. We present a hybrid nested large neighborhood search with variable neighborhood descent algorithm, which is both effective and efficient in solving static complete rebalancing problems for large-scale bike sharing programs.Computational experiments were carried out on the 1 Commodity Pickup and Delivery Traveling Salesman Problem (1-PDTSP) instances used previously in the literature and on three new sets of instances, two (one real-life and one general) based on Share-A-Bull Bikes (SABB) FFBS program recently launched at the Tampa campus of University of South Florida and the other based on Divvy SBBS in Chicago. Computational experiments on the 1-PDTSP instances demonstrate that the proposed algorithm outperforms a tabu search algorithm and is highly competitive with exact algorithms previously reported in the literature for solving static rebalancing problems in SBSS. Computational experiments on the SABB and Divvy instances, demonstrate that the proposed algorithm is able to deal with the increase in scale of the static rebalancing problem pertaining to both FFBS and SBBS, while deriving high-quality solutions in a reasonable amount of CPU time.     

Spatio-temporal analysis of rail station ridership determinants in the built environment

The development of new routes and stations, as well as changes in land use, can have significant impacts on public transit ridership. Thus, transport departments and governments should seek to determine the level and spatio-temporal dependency of these impacts with the aim of adjusting services or improving planning. However, existing studies primarily focus on predicting ridership, and pay relatively little attention to analyzing the determinants of ridership from temporal and spatial perspectives. Consequently, no comprehensive cognition of the spatio-temporal relationship between station ridership and the built environment can be obtained from previous models, which makes them unable to facilitate the optimization of transportation demands and services. To rectify this problem, we have employed a Bayesian negative binomial regression model to identify the significant impact factors associated with entry/exit ridership at different periods of the day. Based on this model, we formulated geographically weighted models to analyze the spatial dependency of these impacts over different periods. The spatio-temporal relationship between station ridership and the built environment was analyzed using data from Beijing. The results reveal that the temporal impacts of most ridership determinants are related to the passenger trip patterns. Furthermore, the spatial impacts correspond with the determinants’ spatial distribution, and the results give some implications on urban and transportation planning. This analysis gives a common analytical framework analyzing impacts of urban characteristics on ridership, and extending researches on how we capture the impacts of urban and other factors on ridership from a comprehensive perspective.

     

Exploring the link between the neighborhood typologies,bicycle infrastructure and commuting cycling over time and the potential impact on commuter GHG emissions

This paper investigates the evolution of urban cycling in Montreal, Canada and its link to both built environment indicators and bicycle infrastructure accessibility. The effect of new cycling infrastructure on transport-related greenhouse gas (GHG) emissions is then explored. More specifically, we aim at investigating how commuting cycling modal share has evolved across neighborhood built-environment typologies and over time in Montreal, Canada. For this purpose, automobile and bicycle trip information from origin–destination surveys for the years 1998, 2003 and 2008 are used. Neighborhood typologies are generated from different built environment indicators (population and employment density, land use diversity, etc.). Furthermore, to represent the commuter mode choice (bicycle vs automobile), a standard binary logit and simultaneous equation modeling approach are adopted to represent the mode choice and the household location. Among other things, we observe an important increase in the likelihood to cycle across built environment types and over time in the study region. In particular, urban and urban-suburb neighborhoods have experienced an important growth over the 10 years, going from a modal split of 2.8–5.3% and 1.4–3.0%, respectively. After controlling for other factors, the model regression analysis also confirms the important increase across years as well as the significant differences of bicycle ridership across neighborhoods. A statistically significant association is also found between the index of bicycle infrastructure accessibility and bike mode choice – an increase of 10% in the accessibility index results in a 3.7% increase in the ridership. Based on the estimated models and in combination with a GHG inventory at the trip level, the potential impact of planned cycling infrastructure is explored using a basic scenario. A reduction of close to 2% in GHG emissions is observed for an increase of 7% in the length of the bicycle network. Results show the important benefits of bicycle infrastructure to reduce commuting automobile usage and GHG emissions.     

Transit trip distribution model considering land use differences between catchment areas

Sustainable land use planning and advanced public transport system are believed to be effective solutions to traffic congestion. To this end, it is important to reveal the relationship between transit ridership and land use. However, current Direct Ridership Models only focus on the relationship between single station's boarding volume and the corresponding catchment area land use. This paper analyzed the ridership distribution between transit stations by considering the land use difference between catchment areas. Land use difference was calculated from point of interest (POI) data extracted from an electronic map of Beijing; transit trip distribution volume was obtained from ‘automatic fare collection’ facility. After data specification, a transit ridership distribution model was proposed and calibrated. The calibration results suggest that land use difference has a directly proportional correlation with transit ridership distribution. The research findings build a bridge between detailed urban form and public transport, which is of significance for the further research of sustainable urban planning. Copyright © 2016 John Wiley & Sons, Ltd.     

A model for the layout of bike stations in public bike‐sharing systems

The problem of designing a layout of bike stations for public bike‐sharing systems entails selecting a number of stations and then constructing them within a planning area having many bike traffic zones and candidate bike stations. In this paper, we proposed a mathematical model to formulate the layout of public bike stations with the objective of minimizing users' total travel time and investment budget constraints. The model can guarantee that the needs for picking up and dropping off bikes amidst all bike travel demands are satisfied. Using this model, the number and locations of bike stations and the number of bikes and parking lockers at each bike station can be simultaneously determined. A typical example solved by lingo solver is created to illustrate the proposed model. Copyright © 2015 John Wiley & Sons, Ltd.     

Cycling to work in 90 large American cities: new evidence on the role of bike paths and lanes

This article analyzes the variation in bike commuting in large American cities, with a focus on assessing the influence of bike paths and lanes, which have been the main approach to increasing cycling in the USA. To examine the role of cycling facilities, we used a newly assembled dataset on the length of bike lanes and paths in 2008 collected directly from 90 of the 100 largest U.S. cities. Pearson’s correlation, bivariate quartile analysis, and two different types of regressions were used to measure the relationship between cycling levels and bikeways, as well as other explanatory and control variables. Ordinary Least Squares and Binary Logit Proportions regressions confirm that cities with a greater supply of bike paths and lanes have significantly higher bike commute rates—even when controlling for land use, climate, socioeconomic factors, gasoline prices, public transport supply, and cycling safety. Standard tests indicate that the models are a good fit, with R ² ranging between 0.60 and 0.65. Computed coefficients have the expected signs for all variables in the various regression models, but not all are statistically significant. Estimated elasticities indicate that both off-street paths and on-street lanes have a similar positive association with bike commute rates in U.S. cities. Our results are consistent with previous research on the importance of separate cycling facilities and provide additional information about the potentially different role of paths vs. lanes. Our analysis also revealed that cities with safer cycling, lower auto ownership, more students, less sprawl, and higher gasoline prices had more cycling to work. By comparison, annual precipitation, the number of cold and hot days, and public transport supply were not statistically significant predictors of bike commuting in large cities.     

Promoting public bike-sharing: A lesson from the unsuccessful Pronto system

In 2014, Seattle implemented its own bike-sharing system, Pronto. However, the system ultimately ceased operation three years later on March 17th, 2017. To learn from this failure, this paper seeks to understand factors that encourage, or discourage, bike-sharing trip generation and attraction at the station level. This paper investigates the effects of land use, roadway design, elevation, bus trips, weather, and temporal factors on three-hour long bike pickups and returns at each docking station. To address temporal autocorrelations and the nonlinear seasonality, the paper implements a generalized additive mixed model (GAMM) that incorporates the joint effects of a time metric and time-varying variables. The paper estimates models on total counts of pickups and returns, as well as pickups categorized by user types and by location. The results clarify that effects of hilly terrain and the rainy weather, two commonly perceived contributors to the failure. Additionally, results suggest that users in the University District, presumably mostly university students, tend to use shared bikes in neighborhoods with a higher household density and a higher percentage of residential land use, and make bike-sharing trips regardless workdays or non-workdays. The paper also contributes to the discussion on the relationship between public transportation service and bike-sharing. In general, users tend to use bike-sharing more at stations that have more scheduled bus trips nearby. However, some bike-sharing users may shift to bus services during peak hours and rainy weather. Several strategies are proposed accordingly to increase bike ridership in the future.     

Influencing factors and heterogeneity in ridership of traditional and app-based taxi systems

The growth of app-based taxi services has disrupted the urban taxi market. It has seen significant demand shift between the traditional and emerging app-based taxi services. This study explores the influencing factors for determining the ridership distribution of taxi services. Considering the spatial, temporal, and modal heterogeneity, we propose a mixture modeling structure of spatial lag and simultaneous equation model. A case study is designed with 6-month trip records of two traditional taxi services and one app-based taxi service in New York City. The case study provides insights on not only the influencing factors for taxi daily ridership but also the appropriate settings for model estimation. In specific, the hypothesis testing demonstrates a method for determining the spatial weight matrix, estimation strategies for heterogeneous spatial and temporal units, and the minimum sample size required for reliable parameter estimates. Moreover, the study identifies that daily ridership is mainly influenced by number of employees, vehicle ownership, density of developed area, density of transit stations, density of parking space, bike-rack density, day of the week, and gasoline price. The empirical analyses are expected to be useful not only for researchers while developing and estimating models of taxi ridership but also for policy makers while understanding interactions between the traditional and emerging app-based taxi services.

     

Bicycling renaissance in North America? An update and re-appraisal of cycling trends and policies

This paper reviews trends in cycling levels, safety, and policies in Canada and the USA over the past two decades. We analyze aggregate data for the two countries as well as city-specific case study data for nine large cities (Chicago, Minneapolis, Montréal, New York, Portland, San Francisco, Toronto, Vancouver, and Washington). Cycling levels have increased in both the USA and Canada, while cyclist fatalities have fallen. There is much spatial variation and socioeconomic inequality in cycling rates. The bike share of work commuters is more than twice as high in Canada as in the USA, and is higher in the western parts of both countries. Cycling is concentrated in central cities, especially near universities and in gentrified neighborhoods near the city center. Almost all the growth in cycling in the USA has been among men between 25-64 years old, while cycling rates have remained steady among women and fallen sharply for children. Cycling rates have risen much faster in the nine case study cities than in their countries as a whole, at least doubling in all the cities since 1990. They have implemented a wide range of infrastructure and programs to promote cycling and increase cycling safety: expanded and improved bike lanes and paths, traffic calming, parking, bike-transit integration, bike sharing, training programs, and promotional events. We describe the specific accomplishments of the nine case study cities, focusing on each city’s innovations and lessons for other cities trying to increase cycling. Portland’s comprehensive package of cycling policies has succeeded in raising cycling levels 6-fold and provides an example that other North American cities can follow.     

Locating rental stations and bikeways in a public bike system

To determine the spatial distribution of rental stations and bikeways in a public bike system, this paper proposes a facility location and network design model. The model is developed as a multi-objective programing problem that considers four objectives (minimizing cyclist risk, maximizing cyclist comfort, minimizing adverse impacts on traffic and maximizing service coverage) and multiple constraints (monetary budget, network connectivity, station spacing, bikeway types, station number and value ranges of decision variables). The ε-constraint method solves the programing problem for the public bike system in Daan District, Taipei City, Taiwan. The nine non-dominated alternatives generated are all markedly better than existing locations of rental stations and bikeways. Scenario analysis results indicate that increasing the construction budget for bikeways significantly improves cyclist safety and comfort whilst increasing the adverse impact on traffic. Planners can use this model to develop public bike systems that spatially integrate rental stations and bikeway networks.     

Impact of weather on urban transit ridership

Utilizing daily ridership data, literature has shown that adverse weather conditions have a negative impact on transit ridership and in turn, result in revenue loss for the transit agencies. This paper extends this discussion by using more detailed hourly ridership data to model the weather effects. For this purpose, the daily and hourly subway ridership from New York City Transit for the years 2010–2011 is utilized. The paper compares the weather impacts on ridership based on day of week and time of day combinations and further demonstrates that the weather’s impact on transit ridership varies based on the time period and location. The separation of ridership models based on time of day provides a deeper understanding of the relationship between trip purpose and weather for transit riders. The paper investigates the role of station characteristics such as weather protection, accessibility, proximity and the connecting bus services by developing models based on station types. The findings indicate substantial differences in the extent to which the daily and hourly models and the individual weather elements are able to explain the ridership variability and travel behavior of transit riders. By utilizing the time of day and station based models, the paper demonstrates the potential sources of weather impact on transit infrastructure, transit service and trip characteristics. The results suggest the development of specific policy measures which can help the transit agencies to mitigate the ridership differences due to adverse weather conditions.     

Exploring the impact of walk–bike infrastructure,safety perception,and built-environment on active transportation mode choice: a random parameter model using New York City commuter data

This study estimates a random parameter (mixed) logit model for active transportation (walk and bicycle) choices for work trips in the New York City (using 2010–2011 Regional Household Travel Survey Data). We explored the effects of traffic safety, walk–bike network facilities, and land use attributes on walk and bicycle mode choice decision in the New York City for home-to-work commute. Applying the flexible econometric structure of random parameter models, we capture the heterogeneity in the decision making process and simulate scenarios considering improvement in walk–bike infrastructure such as sidewalk width and length of bike lane. Our results indicate that increasing sidewalk width, total length of bike lane, and proportion of protected bike lane will increase the likelihood of more people taking active transportation mode This suggests that the local authorities and planning agencies to invest more on building and maintaining the infrastructure for pedestrians. Further, improvement in traffic safety by reducing traffic crashes involving pedestrians and bicyclists, will increase the likelihood of taking active transportation modes. Our results also show positive correlation between number of non-motorized trips by the other family members and the likelihood to choose active transportation mode. The model would be an essential tool to estimate the impact of improving traffic safety and walk–bike infrastructure which will assist in investment decision making.     

Analyzing the impact of a planned transit-oriented development on mode share and traffic conditions

ABSTRACT

Transit-oriented development (TOD) is a popular planning strategy used to maximize accessibility to transit for various trip purposes. The quantitative effects of TOD on travel mode shift and traffic congestion have not been extensively tested in the current literature. This paper utilizes a seemingly unrelated regressions (SUR) mode share model and a mesoscopic dynamic traffic assignment (DTA) model to analyze the impact of a planned TOD in Maryland. The proposed model aims at improving the understanding of the quantitative impacts of such a TOD on mode share and traffic congestion. The main result of the mode share model indicates that the increase in transit ridership for a transit accessible shopping center is not that significant. Local traffic conditions will deteriorate due to a lack of investment in road infrastructure planned for the TOD area. The proposed method could be a valuable tool for other indicative land development or transportation policy analyses.     

Direct and indirect influences on employed adults’ travel in the UK: New insights from the National Travel Survey data 2002–2010

Recent years saw a continuing shift in labour force composition, e.g. greater participation of women and a prominent rise in part-time workers. There are as yet relatively few recent studies that examine systematically the influences on the travel of employed adults from such perspectives, particularly regarding possible transport disadvantages of the fastest growing segments of workers. A robust analysis requires systematic data on a wide range of explanatory variables and multiple travel outcomes including accessibility, mobility and trip frequency for different trip purposes. The UK NTS data does meet the majority of this demanding data requirement, but its full use has so far been hampered by methodological difficulties. To overcome complex endogeneity problems, we develop novel, integrated structural equation models (SEMs) to uncover the influences of latent land use characteristics, indirect influences on car ownership, interactions among trip purposes as well as residents’ self-selection and spatial sorting. This general-purpose method provides a new, systematic decomposition of the influences on travel outcomes, where the effects of each variable can be examined in turn with robust error terms. The new insights underline two direct policy implications. First, it highlights the contributions of land use planning and urban design in restraining travel demand in the 2000s, and their increasing influence over the decade. Secondly, it shows that there may still be a large mobility disadvantage among the fastest growing segments of workers, particularly in dense urban areas. This research further investigates trend breaking influences before and after 2007 through grouped SEM models, as a test of the methodology for producing regular and timely updates regarding the main influences on personal travel from a system level.     

Bike Share: A Synthesis of the Literature

This paper begins by providing an overview of bike share programs, followed by a critical examination of the growing body of literature on these programs. This synthesis of previous works, both peer-reviewed and gray, includes an identification of the current gaps in knowledge related to the impacts of bike sharing programs. This synthesis represents a critically needed evaluation of the current state of global bike share research, in order to better understand, and maximize the effectiveness of current and future programs. Several consistent themes have emerged within the growing body of research on bike share programs. Firstly, the importance bike share members place on convenience and value for money appears paramount in their motivation to sign up and use these programs. Secondly, and somewhat counter intuitively, scheme members are more likely to own and use private bicycles than nonmembers. Thirdly, users demonstrate a greater reluctance to wear helmets than private bicycle riders and helmets have acted as a deterrent in jurisdictions in which helmets are mandatory. Finally, and perhaps most importantly from a sustainable transport perspective, the majority of scheme users are substituting from sustainable modes of transport rather than the car.     

Ridership estimation for short-range transit planning

The objective of this research was to develop a simple transit ridership estimation model system for short-range planning. The main feature of the model system is that it exploits knowledge of transit link volumes which are obtained readily from on-off counts. Extensive use is made of default values for model parameters, taken directly from the transportation literature. The remaining parameters can be derived easily from generally available land-use and socioeconomic data. Expensive household surveys and time-consuming model calibrations are not required. A sequence of simple trip generation, trip distribution and modal split models generate trip-purpose specific transit trip tables, denoted as “trial” trip tables. These trip tables and observed transit link volumes are used in a linear programming model which serves as a correction mechanism. The gain in accuracy is achieved by using the ridership information contained in the transit link volumes. The corrected trip tables may be used in a pivot-point analysis to estimate changes in ridership and revenue. The results of a test application of the model system indicate that it can generate accurate ridership estimates when reliable transit link volumes are available from on-off counts, and when the trial transit trip tables as derived from the first three component models are reasonably accurate.     

Modelling the impact of weather conditions on active transportation travel behaviour

Three weather sensitive models are used to explore the relationship between weather and home-based work trips within the City of Toronto, focusing on active modes of transportation. The data are restricted to non-captive commuters who have the option of selecting among five basic modes of auto driver, auto passenger, transit, bike and walk. Daily trip rates in various weather conditions are assessed. Overall, the results confirm that impact of weather on active modes of transportation is significant enough to deserve attention at the research, data collection and planning levels.