Bike Sharing Research Articles

A Systematic Review of the Coopetition Relationship between Bike-Sharing and Public Transit

The sharing economy, mobile Internet, and smartphones have been widely utilized in recent years to promote the development of bike-sharing services. Bike-sharing serves as a first/last mile travel mode to connect to public transit, which improves trip efficiency, alleviates traffic problems, improves environmental quality, and promotes public health. However, the substitution of public transit by bike-sharing and the decline in public transit ridership have raised concerns among city managers regarding the coopetition between shared mobility services and public transit. To understand the impact of bike-sharing on the decline in public transit and to formulate reasonable synergistic development policies, it is crucial to identify the coopetition relationships between the two. This paper uses a combination of database search and backward snowballing to review existing research. Three research themes were identified: macrolevel studies on bike-sharing and public transit interaction, studies on actual coopetition behaviors based on bike-sharing user surveys, and studies on potential coopetition relationships based on bike-sharing transaction data. The three categories of studies reveal the effect of bike-sharing usage on public transit ridership, the emergency function of bike-sharing in the event of unexpected transit shutdowns, and the substitution and connection relationships between bike-sharing and public transit and the factors influencing them. Finally, this study suggests many directions for future research. This review helps clarify the understanding of the coopetition relationships between bike-sharing and public transit, provides theoretical support to promote the synergistic development of both, and points out ways to deepen the research on the coopetition relationship between the two.

Demand forecasting of shared bicycles based on combined deep learning models

The combined deep learning model for bicycle sharing demand prediction is designed to solve the "last 1 km" problem. At present, there are many companies providing bicycle sharing services at home and abroad, and how to dispatch shared bicycles more efficiently has become an important issue in traffic information research. Sometimes it is difficult to find shared bikes at the exit of some subway stations, along commercial streets, or under some office buildings, while sometimes there are mountains of shared bikes. Therefore, performing demand prediction of shared bikes can efficiently increase the scheduling efficiency of shared bikes, optimize the distribution of shared bikes, and provide more convenient travel services for users. Based on traffic flow prediction theory, this paper studies the spatial and temporal features of shared bicycles. The results show that factors such as time of day, season, weather, and temperature have an effect on the demand for bicycles. Based on the above-mentioned characteristic influencing factors, a CNN-LSTM-Attention algorithm is proposed to forecast the demand for shared bicycles in this paper. Firstly, a CNN-LSTM-Attention model is constructed to predict the demand for bicycle sharing based on the open-source data provided by Capital Bicycle Company. Secondly, it is proved that CNN-LSTM-Attention model is better than 1DCNN-LSTM-Attention, CNN-LSTM, LSTM, SVR-based model and BP neural network model in the precision prediction of shared bicycles, in which the prediction accuracy reaches 97.50%, which confirms the practicality and effectiveness of the model.

Efficient Bike-sharing Repositioning with Cooperative Multi-Agent Deep Reinforcement Learning

As an emerging mobility-on-demand service, bike-sharing system (BSS) has spread all over the world by providing a flexible, cost-efficient, and environment-friendly transportation mode for citizens. Demand-supply unbalance is one of the main challenges in BSS because of the inefficiency of the existing bike repositioning strategy, which reallocates bikes according to a pre-defined periodic schedule without considering the highly dynamic user demands. While reinforcement learning has been used in some repositioning problems for mitigating demand-supply unbalance, there are significant barriers when extending it to BSS due to the dimension curse of action space resulting from the dynamic number of workers and bikes in the city. In this paper, we study these barriers and address them by proposing a novel bike repositioning system, namely BikeBrain, which consists of a demand prediction model and a spatio-temporal bike repositioning algorithm. Specifically, to obtain accurate and real-time usage demand for efficient bike repositioning, we first present a prediction model ST-NetPre, which directly predicts user demand considering the highly dynamic spatio-temporal characteristics. Furthermore, we propose a spatio-temporal cooperative multi-agent reinforcement learning method (ST-CBR) for learning the worker-based bike repositioning strategy in which each worker in BSS is considered an agent. Especially, ST-CBR adopts the centralized learning and decentralized execution way to achieve effective cooperation among large-scale dynamic agents based on Mean Field Reinforcement Learning (MFRL), while avoiding the huge dimension of action space. For dynamic action space, ST-CBR utilizes a SoftMax selector to select the specific action. Meanwhile, for the benefits and costs of agents’ operation, an efficient reward function is designed to seek an optimal control policy considering both immediate and future rewards. Extensive experiments are conducted based on large-scale real-world datasets, and the results have shown significant improvements of our proposed method over several state-of-the-art baselines on the demand-supply gap and operation cost measures.

Exploring the impacts of built environment on bike-sharing trips on weekends: The case of Guangzhou

Shared mobility has brought many disruptive changes to urban transportation systems all over the world. Shared bikes have proven to be among the most successful and influential travel tools in attempting to alleviate the last-mile problem – the difficulty in getting people from transportation stations to their final destinations. This study aims to investigate the impacts of built environment factors on bike-sharing trips. Although many studies have explored these impacts, most have focused on the impacts of urban function, and have paid insufficient attention to the cycling environment. This study used multi-source data, including street view images (SVIs), points of interest (POIs), digital elevation models (DEMs), and road networks, to fully identify the influences of the built environment from five dimensions. The multiscale geographically weighted regression (MGWR) method was used to investigate the impacts of the urban built environment on bike-sharing usage. The results found that high-density roads, recreational POIs, and residential POIs all had positive impacts on the volume of bike-sharing trips in residential areas on weekends, whereas urban greenness negatively impacts bike-sharing usage in parks, because of strict regulations promulgated by local governments. Moreover, the impacts of high-density street networks and residential communities had strong spatial non-stationarity, while the influences of other built environment factors, including road gradient, eye-level greenness, and urban function mixture, were demonstrated spatial stationarity. These findings can facilitate local governments’ and enterprises’ efforts to improve the cycling environment and ensure the efficient management of shared bikes.

A two-stage stochastic programming model for bike-sharing systems with rebalancing

We study the problem of determining the target inventory level of stations in a bike-sharing system, when bikes can be rebalanced later during the day. We propose a two-stage stochastic programming formulation, where the target inventory decisions are made at the first stage, while the recourse decisions, related to rebalancing, are made at the second stage. In the literature, the problem of determining the target inventory levels is solved without taking into account the rebalancing problem, or these two problems are solved sequentially. We prove that more efficient bike-sharing systems can be obtained by integrating these two problems. Moreover, we show that our methodology provides better results than the deterministic formulation, and consider an effective matheuristic, based on the solution of the deterministic problem, to solve the stochastic program. Finally, we compare the solutions obtained by our approach with the actual allocation of bikes in the real bike-sharing system of the city of San Francisco. The results show the effectiveness of our approach also in a realistic setting.

Divide-and-train: A new approach to improve the predictive tasks of bike-sharing systems.

Bike-sharing systems (BSSs) have become commonplace in most cities worldwide as an important part of many smart cities. These systems generate a continuous amount of large data volumes. The effectiveness of these BSS systems depends on making decisions at the proper time. Thus, there is a vital need to build predictive models on the BSS data for the sake of improving the process of decision-making. The overwhelming majority of BSS users register before utilizing the service. Thus, several BSSs have prior knowledge of the user's data, such as age, gender, and other relevant details. Several machine learning and deep learning models, for instance, are used to predict urban flows, trip duration, and other factors. The standard practice for these models is to train on the entire dataset to build a predictive model, whereas the biking patterns of various users are intuitively distinct. For instance, the user's age influences the duration of a trip. This endeavor was motivated by the existence of distinct user patterns. In this work, we proposed divide-and-train, a new method for training predictive models on station-based BSS datasets by dividing the original datasets on the values of a given dataset attribute. Then, the proposed method was validated on different machine learning and deep learning models. All employed models were trained on both the complete and split datasets. The enhancements made to the evaluation metric were then reported. Results demonstrated that the proposed method outperformed the conventional training approach. Specifically, the root mean squared error (RMSE) and mean absolute error (MAE) metrics have shown improvements in both trip duration and distance prediction, with an average accuracy of 85% across the divided sub-datasets for the best performing model, i.e., random forest.

Research on the prediction of bike sharing system’s demand based on linear regression model

This research developed a model of linear regression for forecasting the demand for shared bikes in a bike sharing system. By analyzing a dataset sourced from Kaggle, the study focuses on identifying the factors that have the most impact on bike demand and building a model based on these factors. The methodology involves data cleaning, creating dummy variables for categorical variables, and conducting exploratory data analysis. The features are rescaled, and the model building process includes recursive feature elimination and analysis of VIF and p-values. The outcome indicated that linear regression model accurately predicts bike demand based on various factors. This model can assist employers in adapting their business strategies, understanding customer expectations, and effectively managing bike-sharing systems. The findings contribute to the optimization and success of sustainable urban transportation, emphasizing the potential of bike sharing as an eco-friendly transportation option.

Pricing of Responsive Feeder Transit Considering Competitive Relationships With Bike Sharing

Pricing of Responsive Feeder Transit Considering Competitive Relationships With Bike Sharing

Optimization of bike-sharing repositioning operations: A reactive real-time approach

One of the critical issues in the operation of vehicle-sharing systems is the optimization of the fleet repositioning movements. Repositioning implies the artificial movement of vehicles from places where they accumulate to others in which they are scarce. This yields a higher vehicle availability, without over dimensioning the vehicle fleet and while increasing the vehicle utilization rates. In the particular case of bike-sharing systems, repositioning implies to deploy a fleet of small trucks or vans able to move groups of bicycles from one location to another, with the purpose of maximizing the users’ level of service while minimizing the operating agency costs. This repositioning optimization problem has been previously addressed in the operations research field through Mixed Integer Programing (MIP) and its variants, generally facing two limitations. First, its high computational cost, which prevents achieving direct solutions in realistically large systems. So, it has been necessary to develop heuristics and approximations. And second, its reliance and sensitivity to demand forecasts, with its inherent level of uncertainty. Aiming to overcome these weaknesses, this paper presents a strategy based on a real-time pairwise assignment between repositioning trucks and tasks, in order to optimize the bike-sharing repositioning operations. The proposed method is conceptually simple, less dependent on demand predictions, easily implementable in any coding language and applicable to large systems at a low computational cost. These properties make the method appealing to address the repositioning task assignment in any vehicle-sharing system. On a simulated case study, based on Bicing, the bicycle-sharing system in Barcelona, the proposed strategy has been implemented and compared to the MIP-based routing approach. Results show that the proposed real-time pairwise assignment method is able to significantly improve the performance of the repositioning operations, especially in scenarios where the demand forecast is not accurate. Being based on real-time information, the proposed strategy is flexible enough to solve unpredictable situations. So, the proposed strategy can be implemented as an alternative to MIP-based solutions, or as a complementary strategy for dynamic real-time adaptation of static long-term solutions.

Using open big data sources for customizing bike sharing survey delivering and improving response rate

Convincing respondents to fulfil transport questionnaires is becoming a difficult task. To cope with this challenge, researchers and planners must use all the available resources and the most convenient methodologies. In this study, we evaluated the demand of Madrid's bike-sharing system, BiciMAD, by assessing Open Data from the service operator, to customize the survey plan. Based on this information, we deployed 22 surveyors during a seven-day campaign in the vicinity of the stations. The results of the tailored planning led to achieving a 14% of response ratio, by delivering 7,300 survey cards, versus 9.4% of the massive emailing distribution.

Understanding bikeshare mode as feeder to bus:A case study of Chicago’s Divvy system

The bike-share system gives a decent solution to the first- and last-mile problem. And it connects trip origins/destinations and transit stations. However, current research in this area focuses more on metrobikeshare pattern, and there are still substantial gaps in research in the area of bus-bikeshare pattern. Focusing on Chicago, this study uses Divvy’s location information data to explore the spatial and temporal characteristics of bus-bikeshare transfer trips. Besides, we examine the built environment factors affecting bus-bikeshare transfers. This study uses buffers and time limits to isolate bus-bikeshare transfer trips. Two recognition rules proposed are a maximum transfer time of 20 min and a maximum buffer distance of 100 m. The findings are listed as follows: (1) the average distance and duration of bus-bikeshare trips is 1740.95m and 548.11s; (2) the bus-bikeshare trips show characteristics in different areas, seasons and periods; (3) there is a positive and significant spatial autocorrelation for the bus-bikeshare trips; (4) The bus-bikeshare pattern is influenced by built environment factors in a manner distinct from the bikeshare-only pattern, exhibiting its own unique characteristics.

Enhancing Walkability in Smart Cities: A Comprehensive Literature Review and Strategic Framework

This paper explores the intersection of urban planning and smart city technologies in enhancing walkability, a crucial aspect of creating sustainable and livable urban environments. Through an extensive literature review, the study identifies key factors contributing to walkability, such as safety, accessibility, and comfort, and examines the benefits of walkable cities, including improved public health, economic vitality, and social equity. The concept of the “15-minute city” is introduced as a model for promoting walkability by ensuring access to essential services within a short distance. Additionally, the paper discusses technological advancements and case studies of successful walkability initiatives in smart cities worldwide, highlighting the role of smart walkways, street lighting, public spaces, crosswalks, bike-sharing systems, and digital placemaking in enhancing pedestrian experiences. The comparative analysis of different approaches and strategies provides insights into their key features, benefits, and challenges, offering a comprehensive overview of the current landscape in this field. The study concludes with recommendations for future research and practice to address existing gaps and challenges and further promote walkability in smart cities.

Evolutionary Game Analysis on Sharing Bicycles and Metro Strategies: Impact of Phasing out Subsidies for Bicycle–Metro Integration Model

Evolutionary Game Analysis on Sharing Bicycles and Metro Strategies: Impact of Phasing out Subsidies for Bicycle–Metro Integration Model

AdaBoost.RDT: AdaBoost integrated with Residual-based Decision Tree for Demand Prediction of Bike Sharing Systems under Extreme Demands

AdaBoost.RDT: AdaBoost integrated with Residual-based Decision Tree for Demand Prediction of Bike Sharing Systems under Extreme Demands

Exploring static rebalancing strategies for dockless bicycle sharing systems based on multi-granularity behavioral decision-making

In the continuously evolving context of urbanization, more people flock to cities for job opportunities and an improved quality of life, resulting in undeniable pressure on transportation networks. This leads to severe daily commuting challenges for residents. To mitigate this urban traffic pressure, most cities have adopted urban dockless bicycle sharing systems (UDBSS) as an effective measure. However, making accurate decisions regarding UDBSS demand in different city locations is crucial, as incorrect choices can worsen transportation problems, causing difficulties in finding bicycles or excessive deployments leading to disorderly accumulation. To address this decision-making challenge, it is essential to consider uncertain factors like daily weather, temperature, and workdays. To tackle this effectively, we construct an adjustable multi-granularity (MG) complex intuitionistic fuzzy (CIF) information system using complex intuitionistic fuzzy sets (CIFSs). This system objectively determines classification thresholds using an evaluation-based three-way decision (TWD) method, creating adjustable MG CIF probability rough sets (PRSs). Additionally, to recognize the irrationality of decision-makers (DMs), we propose a method that combines prospect theory (PT) with regret theory (RT), providing a more comprehensive understanding of the influence of DMs' psychological factors on decision outcomes. Building upon these foundations, we present static rebalancing strategies for UDBSS based on MG PRSs and prospect-regret theory (P-RT) within the CIF information system. Finally, using UDBSS data collected from various sensors, we conduct experimental analysis to verify its feasibility and stability. In summary, this approach considers residents’ daily usage preferences, including bicycles utilization and return, with the aim of minimizing unmet resident demands and predicting usage patterns for the next day. It effectively addresses the issue of UDBSS distribution inefficiencies and holds a significant advantage in prediction, making it suitable for broader applications in transportation systems and contributing to the establishment of more advanced modern intelligent transportation systems (MITSs) in the future.

Intention to Utilize Bike Sharing from 2014 to 2022 Using Bibliometric Analysis

Intention to Utilize Bike Sharing from 2014 to 2022 Using Bibliometric Analysis

A GAN-Based Ensemble Model for Predicting the Demand of Shared Bikes in 5G Networks

A GAN-Based Ensemble Model for Predicting the Demand of Shared Bikes in 5G Networks

Designing Bike-Sharing Systems Supported by Data: A Systematic Literature Review

Designing Bike-Sharing Systems Supported by Data: A Systematic Literature Review

Analysis of Changes in Characteristics of Shared Bicycle Trips in Megacities

Analysis of Changes in Characteristics of Shared Bicycle Trips in Megacities

Nonlinear Influence and Interaction Effect on the Imbalance of Metro-Oriented Dockless Bike-Sharing System

Dockless Bike-Sharing (DBS) is an eco-friendly, convenient, and popular form of ride-sharing. Metro-oriented DBS systems have the potential to promote sustainable transportation. However, the availability of DBS near metro stations often suffers from either scarcity or overabundance. To investigate the factors contributing to this imbalance, this paper examines the nonlinear influences and interactions that impact the DBS system near metro stations, with Shenzhen, China serving as a case study. An ensemble learning approach is employed to predict the imbalance state. Then, the machine learning interpretation method (i.e., SHapley Additive exPlanations) is used to quantify the contribution of effects, discover the strength of interactions between factors and uncover their underlying interactive connections. The results indicate the influence of external factors and the relations between pairwise variables (e.g., road density and the day of the week) for each imbalanced state. Provide two quantized sets of factors that can result in the supply-demand imbalance and support future transport planning decisions to enhance the accessibility and sustainability of Metro-oriented DBS systems.