Ying PDF

Evaluating performance of
bicycle sharing system in Wuhan,

China
Zhang, Ying
March, 2011
SUPERVISORS:
1st supervisor: Ir. M.J.G. Brussel
2nd supervisor: Prof. Dr. Huang Zhengdong
Evaluating performance of
bicycle sharing system in Wuhan,
China
Zhang, Ying
Enshcede, The Netherlands, March, 2011
Thesis submitted to the Faculty of Geo-information Science and Earth

Observation of University of Twente in partial fulfillment of the
requirements for the degree of Master of Science in Geo-information
Science and Earth Observation.
Specialization: (Urban Planning and Management)
SUPERVISORS:
1st Supervisor: Ir. M.J.G. Brussel
2nd Supervisor: Prof. Dr. Huang Zhengdong
THESIS ASSESSMENT BOARD

Chairman: Prof. Dr. Ir. M.F.A.M. van Maarseveen
External Examiner: Ing. K.M. van Zuilekom
i
DISCLAIMER
This document describes work undertaken as part of a programme of study at the Faculty of Geo-information
Science and Earth Observation of University of Twente. All views and opinions expressed therein remain the sole
responsibility of the author, and do not necessarily represent those of the Faculty.
ii
ABSTRACT
With the economic development all over the world, there is an increasing usage of automobiles,
which has resulted in a series of urban and transport problems, such as global warming, air
pollution and traffic congestion. Therefore, many countries have taken actions to improve these
problems by encouraging the use of public transportation and green transport. In recent years,
bike sharing system has become a mode of green transport, which has been adopting all over the
world, to be used as a single transport or integrated with other public transport system.
This paper seeks to evaluate performance of bike sharing system in Wuchang district of Wuhan,
in China, based on the existing problems of this bike sharing system. And method on the
combination of GIS and multi criteria evaluation (MCE) was adopted to evaluate this system, by
building a series of indicators which was assumed that would influence the performance of bike
sharing system.
In this study, the research on bike sharing system mainly focused on bike users, bike stations, and
bike lanes. Furthermore, the efficient “performance” of bike sharing system was defined as
operation efficiency and spatial effectiveness of bike sharing system in this research.
The operation efficiency of bike sharing system aimed at understanding the characteristics of
bike users, and evaluating the conditions of bike lanes from the perspective of bike users.
Through the survey on bike users and investigation of bike lanes, a series of the characteristics
of bike users and conditions of bike lanes were understood. While the spatial effectiveness of
bike sharing system aimed at analyzing the characteristics of bike stations, and accessibility
between bike stations and other facilities (e.g. bus stops/subway stations, residential communities,
colleges..). According to the analysis on these two aspects, to find out whether the performance
of bike sharing system is efficient, and to further improve the performance of bike stations in
optimization model.
According to the results of performance evaluation of bike sharing system, a straightforward
method on spatial analysis, which is based on the distribution of bus stops and population, was
adopted to optimize the spatial location of bike stations, this aimed at improving the
interconnection between bike stations and bus stops or residential community.
Keywords:
Bike sharing system, Performance, Evaluation, GIS, Bike stations, Bike users
iii
ACKNOWLEDGEMENTS
I would like to give my gratitude to ITC (Faculty of Geo-Information Science and Earth
Observation, University of Twente) and Wuhan University for providing me this precious
opportunity to study in Netherlands. It broadened my horizon and enriched my knowledge.
Special thanks go to both my supervisors. Firstly, to my first supervisor Ir. M.J.G. Brussel
(University of Twente), for his dedication to always assist me in my thesis and his critical
comments aiming to enhance and improve the present research. Secondly, to my second
supervisor Pro.Dr.Huang Zhengdong (Wuhan University) in China, for his earnest and patient
help and encouragement during the whole study period. His hardworking spirit, intelligent ideas,
critical comments and patience are admirable and have a significant impact on the success of this
research.
I would like to acknowledge Dr. Richard Sliuzas (University of Twente), Dr. Mark Zuidgeest
(University of Twente) and Ing.F.H.M. van den Bosch (University of Twente), for their unlimited
and continuous support and critical comments to my research proposal, and offering advices in
each stage of my research.
I also would like to thank the staff of ITC, for making me have a comfortable and
memorable experience in Enschede.
My sincere appreciation also goes to the ITC friends, Zeng Zheng, Cheng Fangfang, Hao
Pu. The valuable with you all is memorable forever in my life. And to all my course mates: Jowan,
Jiwan, Frinaldi, Mathenge, Parveen, Alex…I say thank you for your support and concern.
To the colleague in Wuhan University, Zhou Jun, Zhang Ning, Meng Xin, Duan Yapeng,
thank you for assisting me in solving the problem encountered in research, and encouraging me
during the difficult period in my study.
Finally, sincere and special appreciation goes to my family, especially my mother, her
warmest love and supports are the most precious wealth I have. Last but not least, I give my
special thanks to beloved boyfriend, for his encouragement, assistance, supporting, and
understanding during the time when I studied in Netherlands and did my thesis.
iv
TABLE OF CONTENTS
ABSTRACT................................................................................................................................................... iii
ACKNOWLEDGEMENTS ...................................................................................................................... iv
TABLE OF CONTENTS ........................................................................................................................... v
LIST OF FIGURES AND TABLES ...................................................................................................... vii
1. Introduction ............................................................................................................................................... 1
1.1 Background and Justification .............................................................................................. 1
1.1.1 Current status of transport development in China ............................................ 1
1.1.2 Bicycle sharing system in China ............................................................................. 1
1.1.3 Importance of developing a bicycle system integrated with public bus ......... 2
1.2 Research Problems ............................................................................................................... 3
1.2.1 Unbalanced supply and demand of bicycle sharing station .............................. 3
1.2.2 Irrational distribution of bicycle stations ............................................................. 3
1.2.3 Insufficient method for performance evaluation ................................................ 4
1.3 Research Objectives and Questions ................................................................................... 4
1.3.1 Research Objectives ................................................................................................. 4
1.4 Research questions................................................................................................................ 4
1.4.1 Conceptual Framework ........................................................................................... 5
1.5 Research Design and Methods ........................................................................................... 5
2. Literature review ........................................................................................................................................ 9
2.1 The development of bicycle sharing system .................................................................... 9
2.1.1 The development of bicycle sharing system in European Countries .............. 9
2.1.2 The development of bicycle sharing system in China......................................11
2.2 Advantage and challenge of developing bicycle sharing system .................................14
2.2.1 Advantages of bike sharing system .....................................................................14
2.2.2 Challenges of bike sharing system.......................................................................15
2.3 Benefits of GIS ...................................................................................................................15
2.4 Service Location Model .....................................................................................................16
2.4.1 Location set covering model & Maximal coverage location model ...............16
2.4.2 Maximal service area model..................................................................................16
2.5 Previous Research on optimizing service location ........................................................17
2.6 Conclusion ...........................................................................................................................17
3. Background of study area ......................................................................................................................18
3.1 General description of Wuhan city..................................................................................18
3.1.1 An profile of study area ........................................................................................18
3.1.2 Description of transportation in Wuhan............................................................18
3.2 Characteristic of bike sharing system in Wuhan ...........................................................20
3.2.1 Introduction of Wuhan’s bike sharing system ...................................................20
3.2.2 Implementation of Wuhan’s bike sharing system .............................................21
3.2.3 Pricing of Wuhan’s bike sharing system .............................................................21
3.2.4 Operations of Wuhan’s bike sharing system ......................................................21
3.2.5 Feedback since operations began .........................................................................22
3.2.6 Future development ...............................................................................................22
4. Data collection and Methodology ........................................................................................................23
v
4.1 Data collection ....................................................................................................................23
4.2 Methodology .......................................................................................................................27
4.2.1 The concept of “performance” on evaluating bike sharing system...............27
4.2.2 Methodology for evaluating “performance” of bike sharing system ............27
4.2.3 Optimization model ...............................................................................................29
5. Analysis and Results ................................................................................................................................31
5.1 Characteristics of bike users .............................................................................................31
5.1.1 Profile of bike users ...............................................................................................31
5.1.2 Bike users’ views to bike sharing system ............................................................32
5.1.3 Conclusion ...............................................................................................................33
5.2 Analysis of bicycle lane......................................................................................................33
5.2.1 Performance evaluation of bicycle lane ..............................................................33
5.2.2 Suggestions for improving the performance of bicycle lane ..........................35
5.3 Spatial analysis on bike stations ........................................................................................35
5.3.1 Characteristics of bike stations ............................................................................35
5.3.2 Accessibility analysis ..............................................................................................39
5.3.3 Discussion ...............................................................................................................46
5.4 Spatial location optimization for bike stations ...............................................................47
6. Conclusions ..............................................................................................................................................52
6.1 Achievement of this research ...........................................................................................52
6.2 Limitation of this research ................................................................................................53
6.3 Recommendations for further research ..........................................................................54
List of Reference .........................................................................................................................................55
vi
LIST OF FIGURES AND TABLES
Figure 1-1 Overview of Chinese cities' bike sharing system .................................................................. 2

Figure 1-2 Bicycle-Sharing System in China ............................................................................................. 2
Figure 1-3 Conceptual Framework ............................................................................................................. 5
Figure 1-4 Research Method ........................................................................................................................ 7
Figure 1-5 Research Design ......................................................................................................................... 8
Figure 2-1 Worldwide Distribution of Bike Sharing system .................................................................. 9
Figure 2-2 Animation of bike sharing system ........................................................................................11
Figure 2-3 BikeGrid (London) ..................................................................................................................11
Figure 2-4 Bike Sharing System in Chinese cities ...................................................................................12
Figure 2-5 Bike Sharing System in Hangzhou ........................................................................................12
Figure 2-6 Financing - Capital Cost..........................................................................................................14
Figure 2-7 Pricing Model in some Chinese cities ...................................................................................14
Figure 3-1 the Location of Wuhan in China and Hubei Province ......................................................18
Figure 3-2 Ownership of Motor Vehicles in Wuhan .............................................................................19
Figure 3-3 Traffic Velocity in Wuhan .......................................................................................................19
Figure 3-4 Wuhan Bikesharing System ....................................................................................................21
Figure 3-5 Bike rental stations in Wuhan (Wuchang) ............................................................................21
Figure 3-6 Manual & Self-service system of Wuhan’s bike sharing system .......................................22
Figure 4-1 Data description .......................................................................................................................23
Figure 4-2 Road Network of width in Wuhan .......................................................................................24
Figure 4-3 Population distribution in Wuhan .........................................................................................24
Figure 4-4 Distribution of bike stations in Wuhan................................................................................25
Figure 4-5 Distribution of bus stops in Wuhan .....................................................................................25
Figure 4-6 Subway stations in Wuhan ......................................................................................................26
Figure 4-7 Land use pattern in Wuhan ....................................................................................................26
Figure 4-8 flowchart of analyzing accessibility between bike stations and specified facilities ........29
Figure 4-9 Flowchart of Coverage Model ...............................................................................................30
Figure 5-1 Bike Users’ age..........................................................................................................................31
Figure 5-2 Bike Users' job ..........................................................................................................................31
Figure 5-3 Bike Users' income...................................................................................................................31
Figure 5-4 Travel Purpose ..........................................................................................................................31
Figure 5-5 Rental Time ...............................................................................................................................32
Figure 5-6 Integrated Travel Mode ...........................................................................................................32
Figure 5-7 the Reason for Renting Bikes .................................................................................................32
Figure 5-8 Problems of Bike Sharing System .........................................................................................32
Figure 5-9 Attitudes of users to bike sharing system ............................................................................32
Figure 5-10 Bike Lane Network ................................................................................................................34
Figure 5-11 Catchment area of bike stations of Wuchang ...................................................................36
Figure 5-12 Bike stations find other stations in certain time stage......................................................37
Figure 5-13 Number of bike stations without finding other stations in certain time ......................37
Figure 5-14 Changes of 8 bike stations from 5 to 10 minutes ............................................................37
Figure 5-15 flowchart of calculating catchment population ................................................................38
Figure 5-16 Distribution of potential catchment population per bike station ..................................39
Figure 5-17 Bus stops containing subway stations .................................................................................39
Figure 5-18 Settings of Layer Properties in service area analysis on bus stops ................................40
Figure 5-19 Distribution of bus stops within bike stations’ coverage ................................................40
vii
Figure 5-20 Number of bus stops covered by per bike station in 10 minutes ..................................41
Figure 5-21 Relationship between bike stations and bus stops (bike station avaliabity) ..................41
Figure 5-22 Distribution of bike stations within coverage of bus stops ...........................................42
Figure 5-23 Number of bike station covered by per bus stop in 10mins ..........................................43
Figure 5-24 Relationship between bike stations and bus stops (bus stop availability) .....................43
Figure 5-25 Distance of allocating residential communities to bike stations within 10 minutes (in
Wuchang) ......................................................................................................................................................44
Figure 5-26 Distance of allocating business to bike stations in 10mins (in Wuchang) ....................44
Figure5-27 Distance of allocating colleges to bike stations in 10mins (in Wuchang) ......................45
Figure5-28 Distance of allocating entertainment to bike stations in 10mins (in Wuchang) ...........45
Figure 5-29 Distance of allocating specific facilities to bike stations in 10 minutes (in Wuchang)46
Figure 5-30 Bus stops which do cover and not cover bike stations in 10mins .................................47
Figure 5-31 Settings of Layer Properties in Service area analysis .......................................................48
Figure 5-32 Service area of bus stops in 10 minutes .............................................................................48
Figure 5-33 Problems of existing bike stations’ locations ....................................................................49
Figure 5-34 Express way of Road network .............................................................................................50
Figure 5-35 Population and bus stops distributed in study area ..........................................................50
Figure 5-36 Optimal bike stations ............................................................................................................51
Figure 5-37 Bus stops and Optimal bike stations ..................................................................................51
Table 5-1 The results of bike station availability in 10mins .................................................................41

Table 5-2 The results of bus stop availability in 10mins .......................................................................42
viii
Evaluating performance of bike sharing system in Wuhan , China
1. INTRODUCTION
With strong economic development all over the world, there is an increasing use of
automobiles, which has resulted in global warming, air pollution, and traffic congestion.
Consequently, many countries are trying to take action to attract more and more people to use
public transport in order to improve these problems, but the accessibility has been a key
determinant that hampers people to choose public transportation as their traffic mode. However,
bicycle was designed with the goal of achieving more efficient transportation, and it is also a
cheaper and green transport mode compared with other modes in developed countries and
developing countries. So there is a trend to combine bicycle with public transportation to attract
more customers, to relieve the traffic pressure and to promote green transport and sustainable
development of transport.
1.1 Background and Justification
1.1.1 Current status of transport development in China

With the rapid development of the economy, urbanization and motorization in China
during the past decades, there is an increase in the number of registered private automobiles (Wei
2010). Since 1978, the urban population has grown from 80 million to more than 560 million, an
annual growth rate of 7.5% (Lin 2002). Vehicle ownership has increased at more than twice this
rate (Cervero et al. 2008). From 2000 to 2003, China’s roads absorbed nearly 14 million additional
vehicles, namely an average of almost 13,000 new cars and trucks per day (Appleyard et al.
2007). According to Feng et al. (2009), studies show that pollutants resulting from vehicle
emission contribute about 60% of the total urban air pollution in China. Next to air pollution,
the urban transport system is also confronting significant challenges in traffic congestion and
energy shortage.
To address these issues, promoting the wide use of public transport can be an effective and
efficient solution. Many Chinese cities have made policies and increased the investment to
promote public transport development. However, public transport development still lags behind
(Li et al. 2010). According to Qiu(2010), public bus trip only accounts for 10% to 25% of the
total residential travel trips in China. According to Li et al.(2010), public bus travel a shorter
distance but take a longer time, especially waiting time and walking time which account for 36%
of the total travel time. And sometimes single-model public transport has a weak accessibility, so
customers would prefer to choose private car or taxi as traffic mode. In order to promote the
public transport development, one of the determinants is to improve the accessibility. In recent
years, many experts think of that it is rational and necessary to establish an integrated urban
transport system, and come up with the idea of “green transport”, for example, establishing
bicycle renting system to induce the public to take public transport.
1.1.2 Bicycle sharing system in China

Cycling is high efficiency, environment friendly, healthy, low pollution, and could help to
reduce congestion, parking needs and energy use; it is regarded as the bright future of the
~1~
transportation system in our future cities. In 1980s, China was named “the kingdom of bicycles”,
though the number of bicycle users has decreased, cycling was and still is one of commuting
modes in most Chinese cities (ITDP-China 2010; Tang et al. 2010).
Because of the increase in traffic and urban problems which have been mentioned earlier,
bicycle sharing systems have been introduced to some Chinese cities to encourage the use of
bicycles which are used as a separate transport or integrated with other public transports, and to
facilitate green transportation. For example, in 2005, China’s first bike sharing system began
operating in Beijing. Within the concern to encourage green transportation, more and more
Chinese cities show their interest to build their own bike sharing systems (Figure 1-1). So far,
there are more than ten cities and city areas that have established bike sharing system
(ITDP-China 2010; Tang et al. 2010) (Figure 1-2).
Figure 1-1 Overview of Chinese cities' bike sharing system
Figure 1-2 Bicycle-Sharing System in China

(Source: http://www.chinabikesharing.org/)
1.1.3 Importance of developing a bicycle system integrated with public bus

Accessibility has been a key determinant related to transport choice. In general, single-model
public transport sometimes have no capability to satisfy customers’ demand that travel from one
place to another place (Grotenhuis et al. 2007), in this case, customers would choose private car
or taxi as their mode to avoid the weak accessibility of public transport. However, using bicycle
as an access or egress trips can substantially reduce the door-to-door travel time of public
transport trips (Martens 2007). Austroads(2005)also proposed that the interaction of bicycles and
buses is more frequently seen to expand the catchments for public transport, when the bicycle is
as a feeder mode.
~2~
Over the past decades, in many industrialized countries, the combination of public transport
and bicycle has been a method to solve a key shortcoming of public transport (Martens 2007),
namely the accessibility of public transport. As a feeder mode, the bicycle is considerably faster
than walking, and more flexible than public bus or other public transports due to its “continuous
character” (that is no travel time gap) which can eliminate the waiting time and scheduling cost
(Keijer et al. 2000). And it offers a number of environmental and social benefits, including
reduction in energy use, air and noise pollution, as well as lower congestion levels and access
routes to bus stops. Last but not least, the combined use of bus and bike may promote the car
free lifestyle, which can improve the development of the green mode of transport.
1.2 Research Problems

The study area is Wuhan, which is a metropolis in the central area of China and
characterized by high population density and large urban scale. Although the public bus service is
the major transit mode in Wuhan, the residents are still suffering traffic congestion, air pollution,
and low quality of public transport service.
In order to relieve the transport problems, a bike sharing system was launched in 2009. This
bike sharing system is built by Wuhan Transport Planning Institute in cooperation with Xinfeida
Bicycle Company, 718 bike rental stations with 20,000 bicycles in eight districts (except Qinshan
district) have been built up since 2009; and these bike rental stations are interconnected with bus
stops, residential communities, large commercial area, colleges, and subway stations. Moreover,
Wuhan Transport Planning Institute and Xinfeida Bike Company have taken some actions to
extend the scale of bike sharing system.
Since the start of bike sharing system, this is always popular with inhabitants. However,
there still exist some problems in the operation of bike sharing system. This is a new and first
bike sharing system in Wuhan, and is still under construction, therefore this research mainly focus
on evaluating bike sharing system based on the probably existing problems of bike sharing
system from bike users’ perspective in the reality.
1.2.1 Unbalanced supply and demand of bicycle sharing station

Supply and demand can be one of the important determinants that influence the quality of
a bicycle sharing system. A balanced supply and demand can not only enhance utilization of
bicycles, but also promote the development of the bicycle sharing system. According to Tang et
al. (2010), unbalanced volumes will bring about a redistribution problem, which will cost a big
amount of money and become the largest problem of the bike sharing system, especially in the
peak hours when it is hard to find a bike in the hot spots and it is no room to return bikes. In this
case, it definitely reduces the recycle use of bikes and decreases the quality of bicycle sharing
system.
1.2.2 Irrational distribution of bicycle stations

Accessibility is an advantage of bike sharing system, and is deeply impacted by the
distribution of bike rental stations. MacIntyre (2006) proposed that NMT users are sensitive to
travel distance. With respect to the performance of the bike sharing system, which as an access
or egress trips in the integrated public transport system, are impacted by the efficiency of
transport network. According to Zuidgeest et al. (2009), the distribution of bicycle stations
influences the efficiency of the bus-bike integrated system, when the access distance is short it is
unlikely that users bike instead of walk to the facilitated bus stop, however , when the access
distance is longer than a threshold it is likely that people choose another access mode instead of a
bike. In order to enhance the accessibility and utilization of the bicycle system, a rational and
~3~
efficient distribution of bike rental stations should be well integrated with bus stops and residents,
that is to say, it has a suitable travel distance and travel time that contribute to a trip.
1.2.3 Insufficient method for performance evaluation

In China, local public transport plans are made based on policies that are made by
MOHURD (Ministry of Housing and Urban-Rural Development of the People’s Republic of
China ) cited by Wei (2010), which regulate the public bus system that including route length,
route density, average transfer coefficient and net transit capacity. In the conventional planning,
planners plan and design transport network just based on policies and qualitative analysis, but
scarcely adopt quantitative methods and technical skills to analyze traffic demand and service
coverage, which also results in traffic congestion and unbalanced demand and supply in real
situation of transport.
In Wuhan, the distribution of bike rental stations are just based on the investigation of
different institutes, which tended to distribute stations around bus stops, residential communities,
large-scale commercial districts, colleges, metro stations (XFD 2010). There is no scientific
analysis of traffic demand and coverage of bike rental stations based on GIS techniques. So, in
the implementation of bicycle sharing system, there are existing problems of irrational
distribution and insufficient capacity of bike rental stations. It is sensible to apply GIS techniques
in the evaluation and optimization of bike sharing system, instead of single qualitative analysis.
Thus, this study mainly focuses on evaluating the performance of bike sharing system,
based on a series of qualitative and quantitative methods and indicators, for example, the
characteristics of bike stations, and the accessibility of bike stations. In addition, GIS techniques
will be applied in the evaluation and optimization process.
1.3 Research Objectives and Questions
1.3.1 Research Objectives

Main objective:
z To evaluate whether the performance of bicycle sharing system in Wuhan is efficient.
Sub-objectives:
z To understand the characteristics of bicycle users;
z To select and perform indicators for the evaluation of bike lane;
z To analyze characteristics of bike stations;
z To analyze accessibility between bike stations and other facilities;
z To establish and implement a model for spatial location optimization;
1.4 Research questions

z To understand the characteristics of bicycle users;
Q1: what elements will be set as “characteristics” of bike users?
z To analyze characteristics of bike stations;

Q2: How to define and perform “characteristics” of bike stations for analysis?
z To analyze accessibility between bike stations and other facilities;

Q3: What method will be implemented for accessibility analysis?
z To select and perform indicators for the evaluation of bike lane;

~4~
Q4: what criterions can be set for evaluation?
z To establish and implement a model for spatial location optimization;

Q5: Which model will be set for spatial location optimization?
1.4.1 Conceptual Framework

The benefit of a bicycle system is that it can offer door-to-door service, so the combined
bus-bike system can enhance the accessibility of public transportation trips. However, the
efficiency of bike sharing system’ performance substantially determines the value of bike sharing
system. In this research, the efficiency of bike sharing system’s performance is defined as
operational efficiency and spatial effectiveness of bike sharing system, and this will be introduced
in more detail in section4.2.1. Therefore, the performance evaluation of bike sharing system will
be operated from three aspects: first, to analyze the characteristics of bike stations; second, to
analyze the level of accessibility between bike stations and other facilities including bus
stops/subway stations, residential communities, colleges, and business area; third, to analyze bike
lane network in study area. This research seeks to detect the performance of bike sharing system
and improve the integration of public bus and bike sharing system. (Figure 1-3)
Figure 1-3 Conceptual Framework
1.5 Research Design and Methods

The main purpose of this study is to evaluate the efficiency of bicycle sharing system based
on spatial analysis, which uses the bicycle sharing system in Wuchang district of Wuhan as a case
study and optimize bike stations with poor efficiency. Figure 1-4 outlines the methods proposed
for answering specific questions under each of the sub-objectives identified in Section 1.3.1.
Figure 1-5 outlines the operational plan under which the whole study will take place. It
incorporates five stages, including field work and data integration (stage1), analysis of
~5~
characteristic of bike stations (stage2), accessibility analysis of bike stations (stage3), evaluation
of bike lane (stage4), station optimization (stage5).
x Understanding the characteristics of bicycle users

Understanding characteristics of bicycle users aims to investigate bike users’ profile and
attitudes to bike sharing system, and this can be related to location and capacity of bicycle rental
stations. With respect to the investigation of bike users’ profile, it can be including travel mode
and destinations of bike users, required time and cost of bike users, and so on. If there is no
available data about the characteristics of bike users, it is necessary to do a survey in a small-scale
area.
x Analysis of the characteristics of bike stations

The characteristics of bike stations incorporate two parts: spatial characteristic of bike
stations, and non-spatial characteristic of bike stations. With respect to the spatial characteristic
of bike stations, catchment area of bike stations in a certain time, and station spacing between
bike stations will be analyzed. And for analyzing the non-spatial characteristic of bike stations,
potential catchment population of bike stations will be calculated. Both of these indicators can
reflect the characteristics of bike stations. After analysis, catchment area, station spacing and
catchment population of bike stations can be obtained, which can be useful and necessary for the
further analysis.
x Accessibility analysis of bike stations

In this stage, accessibility analysis is divided into two parts which are accessibility analysis
between bike stations and specific facilities, and accessibility analysis between bike stations and
bus stops or residential communities. Indicators which influence accessibility and methods for
accessibility analysis can be derived from literature review. After accessibility analysis, bike
stations with weak accessibility can be obtained.
x Performance evaluation of bike lane

The performance evaluation of bike lane aims to evaluate whether the performance of bike
lane is efficient, which can be an important determinant that influence the performance of bike
sharing system. And the performance of bike lane will include two aspects: the construction of
bike lane network, and the operation of bike lane. Through analyzing the two aspects of bike
lane, find out whether the bike lane is efficient and whether there is weak point of bike lane.
x Station optimization
Indicators and constraints for optimizing spatial locations of bike stations can be obtained
from literature review. Data that used for station optimization can be prepared in the previous
work. Selected station with poor performance will be upgraded related to characteristics and
accessibility analysis of bike stations. A useful method for spatial location optimization will be
selected and implemented.
~6~
*Note: FD: Field work and Data analysis; CB: Characteristic of bike stations; AA: Accessibility Analysis; BL: Evaluation of bike lane;
SO: Station Optimization.
Figure 1-4 Research Method
~7~
Data collection and integration Data requirement
Users’ profile
Understand characteristics of
potential bike users Travel purpose
Create database Stage 1
Analyze the coverage of bike

Bike station
stations
Road network
Analyze station spacing between
bike stations
Literature
Bike station
Calculate the potential population of
bike stations Population
Analyze the catchment population Potential population

Of bike stations Coverage
Stage 2
Review
Identify indicators for accessibility Road network

analysis
Bike station
Accessibility analysis of bike stations
to specific facility Facility location
Bus stop
Accessibility analysis of bike stations
to bus stop and residential Community
Stage 3
Evaluating the construction and

operation of bike lane network Road network
Suggestions for improving the Bike station

performance of bike lane
Stage 4
Select bike station with poor Performance of

efficiency Bike stations
Set indicators for optimizing
Spatial location optimization

Stage 5
Figure 1-5 Research Design
~8~
2. LITERATURE REVIEW
2.1 The development of bicycle sharing system
Bike sharing systems have been appearing in more and more cities around the world in the
last few years, which have received increasing attention in recent years with initiatives to increase
bike usage and improve the last mile connection to other modes of transit and lessen the
environmental impacts of our transport activities(DeMaio 2009). According to The Bike-sharing
Blog(2009), based on an unscientific count of the bike sharing services listed on the bike-sharing
world map, there was approximately 160 bike sharing systems globally at the end of 2009, which
is up nearly 74% from 92 bike sharing systems at the end of 2008. Bike renting systems contain a
large fleet distributed at high and medium density areas, and usually allow for one way trips and
sometimes provide short-term access(PBOT 2010).Bicycles from a bike renting station are
typically available in cities with other forms of public transportation to provide for intermodal
transportation options for the public(UtilityCycling 2008). Users check out a bike by using credit
card, membership card, or cell phone at a docking station.
As the development of better methods of tracking bikes with improved technology, this
give birth to the rapid expansion of bicycle renting systems throughout the Europe and most
other continents during the recent decades(DeMaio 2009).
According to Bührmann(2008), public bicycle systems are not only for free, but have a high
added value in the long run if properly implemented. They can help to come to a real “bicycle
culture” and to change people’s travel behavior.
Figure 2-1 Worldwide Distribution of Bike Sharing system
2.1.1 The development of bicycle sharing system in European Countries
2.1.1.1 History of European bike sharing system

Bike sharing system originated in Europe(UtilityCycling 2008).According to DeMaio(2008),
there have been three generations of bike renting system during the past 40 years. The 1st
generation began in 1964 in Amsterdam with the white bikes. The operators gathered a handful
of bicycles, which they painted white and distributed around the city Amsterdam for providing
~9~
bikes to public for free(UtilityCycling 2008). But things didn’t go as planned, due to the bikes
were thrown into canals or appropriate for private use, so this system collapsed within
days(DeMaio 2008).while a strong memory or urban legend still lingers around the idea of the
white bicycle in Amsterdam(UtilityCycling 2008).
Nearly 30 years later, in 1995, the 2nd generation was operated in Copenhagen called
Bycyklen, with many improvements over the previous generation. That bike sharing system were
specially designed, users could pick up and return bikes at specific locations throughout the
central city areas with a coin deposit. While more formalized than the previous generation-with
stations and organization to operate the program – these bikes still experienced theft due to the
anonymity of the customer(DeMaio 2008).
The 3rd generation system is smartened with a variety of technological improvements
including electronically locking rack or bike locks, telecommunication systems, smartcards,
mobile phone access and on-board computers. Bike sharing system grew slowly in the following
years until the launch of Velo’v in Lyon, France(DeMaio 2008).
In the past years, many cities in European countries have been successful in the project of
bicycle renting system. According to PBOT(2010), the Velib system in Paris is generally
considered successful. In two years, Velib’s 20,000 bicycles generated over 54 million trips. On
average, Parisians and its visitors generate 74,000 bike sharing trips a day. The Velo’v program in
Lyon is arguably more successful with an average of 22,000 trips on a 3,000 bike fleet – or
roughly seven trips per bike per day. And Barcelona’s Bicing project planners projected to gain
2,000 bikes sharing member per month. After just four months, 82,000 people had joined (source:
Clear Channel Outdoor). Barcelona has doubled the size of its program to 6000 bikes(PBOT
2010).
In recent years, Europe has been a real boom of public bicycle schemes, which make the
bicycle accessible as part of the public transport system. The number of cities that already
implemented or plan to do so, increased strongly. And the scale of the systems has also reached
new heights. Beyond Europe, the interest in the concept of public bicycles is also rising, e.g. in
the US, in Canada, Australia, Argentina, or Israel.(Bührmann 2008)
2.1.1.2 European approaches to bike sharing system planning

The bicycle sharing system in European cities has already been mature in some aspects,
compared to the system in Chinese cities. According to WANG et al.(2009), most of the
European bike sharing system are public-private-partnership, and the users are required to
register it by paying membership fee and members can use the bicycles free of charge within half
an hour. And many modern techniques are applied in the 3rd generation of European bike
sharing system, for example, network techniques, and wireless communication techniques etc.
With respect to the bike sharing system in some of the European cities, the bike users can
obtain the data from the operator’ websites and observe the dynamic change of capacity in each
bike station over 48 hours on operator’ websites, which are all in JavaScript (Figure 2-2). To a large
extent, this is great convenient for the users(Oliver 2010).
~ 10 ~
Figure 2-2 Animation of bike sharing system

The planner of bike sharing system in London also developed a visible technique –
BikeGrid which is a very sophisticated and data-rich visualization that provides a detailed
overview of the actual performance of the Barclay Cycle Hire scheme. It shows the real-time and
historical availability of the “Boris bikes”, which are located in over 300 docking stations spread
around central London. On this dynamic websites, the initial map view shows all docking stations
are rectangles, sized by the capacity of the station. Darker stations are full, while lighter stations
are empty. A unique “grid view” reveals the status of each docking station more clearly: the
proportion of bikes at each station is indicated by the height of the blue bar. Very full or very
empty stations are shown with a lighter or darker border respectively. In addition, the actual
status of each bike station over the last 24 hours can be seen as a line graph, highlighting the
urban dynamics of bikes going in and out the city centre over a time span of a typical
day(Barclays 2010).
Figure 2-3 BikeGrid (London)
2.1.2 The development of bicycle sharing system in China
2.1.2.1 Overview of Chinese bike sharing system

Bicycle sharing system has also become popular in China, which is a number of bicycles
made available for shared use amongst individuals who do not own the bicycles(Schroeder 2010;
WIKIPEDIA 2010). According to ITDP-China cited by Schroeder (2010), over 60% of people
use public bike because it is more efficient than the transport they used earlier.
In China, the first bike sharing system was launched in Beijing in 2005, it was not a big
system compared with such a big city and without governmental involvement. And there was no
~ 11 ~
digital information technology used in this system(ITDP-China 2010; Tang et al. 2010). In
addition, there are other cities also show there interest in building bike sharing system, such as
Shanghai, Hangzhou, Guangzhou, Wuhan, and so on; moreover, the bike sharing system in some
of these cities are working well(Figure 2-4).
Figure 2-4 Bike Sharing System in Chinese cities
Bike sharing system was started in Hangzhou in May 2008, which is operated by Hangzhou
Public Transport Bicycle Service Development Co; and 2,204 bike rental stations with 50,000
bikes are built. Moreover, some bike stations are open for 24 hours, and citizens can use bikes in
one hour without charge. Hangzhou bike sharing system has been the most successful scheme in
China, where the bike rental stations can be found every 100 meters, compared to the 300 meters
in Paris. And during the first year operation, no bikes were stolen and few bikes were broken or
vandalized compared to the half of that were stolen or broken in Paris(ITDP-China 2010;
WIKIPEDIA 2010).
Figure 2-5 Bike Sharing System in Hangzhou
In order to cater for the 2010 WorldExpo, Shanghai has also launched a bike sharing
program in March 2009, which is operated by Shanghai Forever Bicycle Company. And bike
rental stations are open for 24 hours. This bike sharing system can be accessible by RFID cards.
Users can purchase 100 ride credits for about 30 dollars. Short rides are rewarded credits and
~ 12 ~
longer rides subtract credits once the bike has been re-docked. Shanghai plans to expand to 3,500
bicycle hot spots throughout the entire city by 2010(ITDP-China 2010; WIKIPEDIA 2010).
Guangzhou launched the bike sharing system in June 22nd 2010, which is initiated by
Guangzhou government and operated by Guangzhou Public Bicycle Operation and Co., and
ITDP (Institute for Transportation and Development Policy) was asked to provide assistance
with issues such as station distribution, hardware technical support and operational
recommendations. This bike sharing project contains 3 phases, providing 15,000 bicycles, to
along BRT corridor and nearby living and business area, and to set stations around bus stops and
in Jinshazhou and Daxuecheng. The bike sharing project adapt automatic charging system by
using Yangchengtong IC card (can also be used on metro and bus) or a Public Bicycle Service
card to activate either manual or self service, and citizens can use public bike free of charge
within one hour. Since the first day of starting, the public bike project has been welcome by the
citizens and media. In only 2 months, the number of rental per day has reached 5000 and keeps
growing(Warrier 2010).
In addition, Wuhan is the first city that launches a free public bike sharing system in China
since 2009. Till now, more and more Chinese cities start to build a public bike sharing system,
such as Shenzhen, Suzhou, Chengdu, and so on. The bike sharing system has been widely
spreading to more and more Chinese cities.
2.1.2.2 The profile of Chinese bike sharing system

According to Schroeder(2010), the evidence derived from bike sharing system in Beijing,
Shanghai and Hangzhou shows that public bike is public transit, of which vast majority trips are
commuter and utilitarian trips. And public bike also improves the usage of other public
transports, when it is combined with other public transports. Furthermore, public bike is faster
than bus or rail for short trips and flexible route, and runs 24 hours in some Chinese cities.
z Technology system
In China, there are differences in technology system of bicycle sharing system in different
cities. That is to say, there are high technology system and low technology system. For the bike
sharing system which is of high technology, it needs higher initial capital costs and lower
maintenance cost, but it is easier to integrate with other transports (e.g. bus, metro, car paring).
While in relation to the bike sharing system which is of low technology, it needs lower initial
capital costs, but needs higher operating costs and to create employment due to there is no auto
locking mechanism, and to distribute more bikes per square meter of pavement space(Schroeder
2010).
To some extent, the low technological bike sharing system is inconvenient for users to rent
bikes and also results in some problems. For example, bike stations without good quality docking
stations so that the bicycles look disorderly and messy, and there is long waits for service due to
the system is not automated(Schroeder 2010).Therefore, it is necessary to improve the
construction and management of bike sharing system with low technology system.
z Financing
In some Chinese cities, there are two approaches to deal with the financing of bike sharing
system. One is that capital costs are usually absorbed by the government. The other one is to
gather money from operational revenue which includes system sponsorship, advertising revenue,
annual membership fee, usage fee, retail revenue, etc (Schroeder 2010).
~ 13 ~
(Source: http://www.slideshare.net/rgadgi/guangzhou-bike-share-nitin-warrier)
Figure 2-6 Financing - Capital Cost
z Operation
In China’s bike sharing system, a public bicycle card is provided to citizens, which also can
be used on metro and bus in some Chinese cities, to rent bikes. And bike users need to put a
deposit into public bicycle card to activate it which is used for renting and returning bikes in
either manual or self service. In some Chinese cities, the temporary card is provided to people
who want to use public bike temporarily.
In most of the Chinese cities, bikes can be borrowed and returned at any bike station in
opening time, and 24 hours are available in some cities or stations(ITDP-China 2010).
z Pricing
Apart from the free bike sharing system in Wuhan and credits system in the bike sharing
system of Shanghai and Beijing, most of the Chinese cities adopts a same pricing model for
using bikes in bike sharing system. That is to say, users can rent bikes without charge in the first
one hour, but need to pay for renting bikes in additional hours with certain money. The goal of
the pricing model is to encourage short trips and promote the recycle use of bikes, instead of
maximizing revenue. For example, in Hangzhou, 95% of the trips are less than 1 hour, so they
are free(Schroeder 2010).
Figure 2-7 Pricing Model in some Chinese cities

z Marketing and maps
The website and marketing is a platform to inform users and keep them up to date on new
developments in the bike sharing system(Schroeder 2010).
2.2 Advantage and challenge of developing bicycle sharing system
2.2.1 Advantages of bike sharing system

Bicycles have several advantages as a mode of transportation for short-distance urban trips:
they reach underserved destinations, require fewer infrastructures and generally do not add to
congestion. In addition, they are relatively inexpensive to purchase and maintenance, do not
create pollution in their operation, and provide the user with exercise. Their value is undeniable
when one also considers that these bicycles may increase trips on other modes of public
transportation by expanding the reach of trains and buses(DeMaio 2008).
The implementation of a public bicycle scheme brought about many benefits. It provides a
fast, convenient and flexible inner urban transport option, and can be a “door opener” to
increase the acceptance of cycling as urban transport mode in cities which still lack a good level
~ 14 ~
of bicycle use. It also makes sense in cities that have a good level of cycling as it increases
sustainable mobility choices at low cost when compared to other public transport measures.
Furthermore, it encourages intermodal travelling(Bührmann et al. 2008).
According to PBOT(2010), the bike sharing system has the potential to introduce new
people to urban cycling, reduce peak-hour pressure on transit, reduce automobile trips, and
improve livability. And achieve a change towards a more sustainable multimodal travel behavior
(“the right mode for the right trip”). If the systems are properly implemented, they can be part
of the “bigger puzzle” of an integrated urban transport strategy, which enables cities to reduce
motorized traffic and its environmental impact(Bührmann 2008).
2.2.2 Challenges of bike sharing system

According to PBOT (PBOT 2010), the bike sharing system have also faced a few of
challenges, that is theft and vandalism, maintenance of bicycles and docking stations, helmet use
and cost. Some of these challenges are also emerging in China, like vandalism, maintenance of
bikes and docking stations and cost.
There are also other manifold challenges for the implementation of a public bicycle system.
First, getting it started is not as easy as it seems (need for integrated approach, infrastructure,
scale of scheme and layout, traffic safety etc.). Cities need advice to avoid mistake. Second,
financing model is a key aspect. There are many options, which should be thoroughly evaluated
by a city before tendering and contracting a service. For example, outdoor advertisement contract
(e.g. Rennes, Lyon); Service paid through parking revenues (e.g. Barcelona, 10 years – 22,3Mio.);
advertisements on bicycles (e.g. OYbike, Next); and others (e.g. backed-up by operator).
Furthermore, automatic systems are often not very suitable for small and medium cities,
which are well known the bicycles as urban transport mode. Last, achieving real long term impact
needs continuous development of urban transport strategies. The direct impact bicycle schemes
regarding the reduction of motorized traffic in the city centers and of CO 2 emissions is limited.
In the mid-to long run, however, public bicycles can be an important building block towards a
multi-modal and more sustainable travel behavior(using the right mode for the right
trip)(Bührmann 2008).
2.3 Benefits of GIS

Geographic Information Systems (GIS) were initially developed as tools for the storage,
retrieval and display of geographic information, which is capable of the geographic analysis of
spatial data which were either poor or lacking in these early systems. Moreover, the potential of
GIS to become a powerful tool for spatial analysis in general and for spatial data analysis in
particular is by now commonly accepted in the research literature. Following the needs for better
integration of GIS and the methods of spatial analysis, various alternatives have now been
suggested(Anselin 1993; Fotheringham et al. 1994).
A capability for GIS to integrate digital maps and spatial analysis has made it a powerful tool
for public transit planning. Parallel with the wide adoption of GIS technology in the
transportation has grown rapidly, which dates from the earliest interest in GIS in 1960. With
respect to the application of GIS in transportation, a wide variety of researches have been done,
ranging from transit route selection to market potential evaluation to potential bus stop
placement to transit GIS data model design(Goodchild 2000; Lao et al. 2009)
The reason that GIS should be used in public transportation planning can be seen from two
perspectives. The one perspective is using a GIS for different spatial analysis; a quite common
opinion is that these systems often lack sufficient tools for studying mobility. Therefore, the aim
of this perspective is to include such modeling tools in an existing GIS. Another perspective is
that of transport modelers wanting to use a GIS as an aid to visualize modeling results or to
~ 15 ~
input data to such a tool; and the aim of this perspective is to simply transform data from
planning tools to a GIS, and vice versa(Hall 2006).
With respect to the non-motorized travel, GIS can be used to enhance bicycle demand
forecasting and facility analysis by permitting spatially-based analysis, otherwise, which might be
difficult or impossible. Moreover, GIS can also be used to display and communicate information
relevant to bicycle planning(U.S.DTFHA 1999). In this study, GIS based spatial analytical method
would be used to evaluate the performance of bike sharing system.
2.4 Service Location Model
2.4.1 Location set covering model & Maximal coverage location model
Coverage of location has been an aspect of influencing service location. Many methods
have been developed to solve coverage problems. Location set covering problem (LSCP)
proposed in Toregas et al. cited by (Murray 2001), and Maximal coverage location problem
(MCLP) propose in Church and ReVelle (1974) cited by (Murray 2003) are two common models
used for analyzing stop placement efficiency.
z Location set covering problem
The LSCP was originally utilized for locating a minimum number of emergency service
facilities and has also been suggested for identifying a minimum number of express bus stop
locations. The use of the LSCP for assessing redundancy in service coverage for an existing
public transportation system does not appear to be a previous application area(Murray 2001). In
the LSCP, the optimum number of facilities is one aspect of the solution to the problem, and the
constraint requires for all demands must be covered by at least one facility(cited by MAHMUD et
al.(2009)).
z Maximal coverage location problem
The MCLP was proposed to seek the maximum population that can be served by a limited
number of facilities within a stated service distance or time(MAHMUD et al. 2009). In the MCLP,
the number of facilities is known a priori and the objective becomes to maximize services for
demands (cited by MAHMUD et al.(2009)).
2.4.2 Maximal service area model

Conventional facility location models only define a facility’s service area simply as a
circular-shaped region based on a specified radius. Such definition might be appropriate for
facilities which are not influenced by topographical barriers. But for bike station, accessibility is
an important requirement. Therefore, road accessibility should be taken into account in
emergency facility (e.g. bike rental facility) location problem to improve emergency services (e.g.
bike renting service). GIS can serve this requirement through network analysis. Network analysis
in GIS takes into account network attributes such as road width, speed limit, barriers, turn
restriction and one way restriction. This advantage provided by GIS should be incorporated in
the service area calculation to obtain a more realistic model(MAHMUD et al. 2009).
The maximal service area problem (MSAP) is created as a modification of the MCLP which
makes use of the capability of GIS to generate service areas of facilities as travel time zones. In
the travel time zones, service area polygons are generated based on the network through a
network analysis in a vector GIS environment. Network analysis in GIS takes into account
network attributes, such as road width, speed limit, barriers, turn restriction and one way
restriction. The MSAP does not take into consideration capacity constraints of facilities in
calculating the service area. It is designed as a discrete model where a specified number of facility
sites that achieves the best objective function value of the problem are selected out of a finite set
~ 16 ~
of potential sites. The actual area of coverage can be calculated in the GIS once the set of facility
sites has been selected.
2.5 Previous Research on optimizing service location

Service location has played an important role in developing and planning a multimodal
transportation system (Murray 2001; Rybarczyk et al. 2010). And many methods on optimizing
and planning service location have been proposed. Murray(2001) proposed a model that supports
strategic analysis of service coverage inefficiency and stop spacing optimization along a route to
address strategic aspects of service access. In Chien et al.(2002)’s study, a mathematical model
was developed to improve the accessibility of a bus service by optimizing the number and
locations of bus stops. MAHMUD et al.(2009) also present and discuss the new developed
model to maximize total service area of a fixed number of facilities. In order to reduce
redundancy bus stops, a muticriterion buffer analysis was performed using ArcGIS to identify
suitable locations for bus stop, considering the parameters of distance, importance of location,
willingness to walk and junction points (Alterkawi 2006).Ziari et al.(2007) concluded that past
research on optimizing location of bus stops can be classified into three categories.
Apart from methods on optimizing bus stops, many methods on bicycle facility planning
also have been proposed. According to Rybarczyk and Wu (2010), typical bicycle facility planning
models can be divided into two groups: supply-and demand-based models. And supply-based
bicycle facility planning relies on two overarching theories: (a) all major arterials and collectors
should have bike facilities, or (b) a quantitative model, such as bicycle level of service, hazard
score analysis, or bicycle compatibility index should be calculated prior to bicycle route planning.
Service location models concern the provision of a service to satisfy a spatially dispersed
demand. A demand for the service exists at a large number of widely dispersed sites, it is
impossible to provide the service anywhere(MAHMUD et al. 2009). It is important to apply a
suitable method in the research of a specific study area.
2.6 Conclusion
Because of the increasing environmental issues and transport problems, bike sharing system
has become a reigning project in all over the word. And the overall bike sharing systems in
Chinese cities are in building and improving phrase.
In this research, GIS technique is applied in analysis. And service location model –
especially the maximal service area model- is used for analyzing and optimizing bike stations.
Moreover, optimization of bicycle station mainly focuses on the efficiency of service coverage
and station spacing of bike stations.
~ 17 ~
3. BACKGROUND OF STUDY AREA

3.1 General description of Wuhan city
3.1.1 An profile of study area

The study area is urban area of Wuchang district in Wuhan city. Wuhan is the capital of
Hubei province, China, and is the most populous city in central China. It is situated in the middle
of Hubei Province, at the east of Jianghan Plain, and the confluence of the middle reaches of
the Yangtze River and Hanshui River. The metropolitan area comprises three parts – Wuchang,
Hankou, and Hanyang, commonly called the “Three Towns of Wuhan”. The consolidation of
these three cities occurred in 1927 and Wuhan was thereby established. These three parts face
each other across the rivers and are linked by bridges, including one of the first modern bridges
in China, known as the “First Bridge”. It is simple in geographical structure – low and flat in the
middle and hilly in the south, with the Yangtze and Han rivers winding through the city.
Arising out of the conglomeration of three boroughs, Wuchang, Hankou, and Hanyang,
Wuhan is known as “the nine provinces’ leading thoroughfare”; it is a major transportation hub,
with dozens of railways, roads and expressways passing through the city. The city of Wuhan, first
termed as such in 1927, has a population of approximately 9,100,00 people(2006), with about
6,100,000 residents in its urban area. Now Wuhan is recognized as the political, economic,
financial, culture, educational and transportation center of central China.
a. Wuhan in China b. Wuhan within Hubei Province

Figure 3-1 the Location of Wuhan in China and Hubei Province
3.1.2 Description of transportation in Wuhan
3.1.2.1 Current status of transportation in Wuhan

Although the speed of constructing major infrastructures has been accelerated and the
integrated transportation system has been improving, the inhabitants are still experiencing that
electric bicycles are faster than cars in urban area of Wuhan(WHTPI 2010).
Nowadays, the car owners are consistently increasing at the speed of more than 10,000
vehicles per month in Wuhan. In 2009, the number of private cars has been up to 402,000.
Therefore, the travel demand is increasing quickly, and the traffic congestion in central urban area
is still serious and has been spread to the peripheral area. In 2007, during the peak hours, there
~ 18 ~
are 48 intersections where the traffic volume is more than 5,000 vehicles per hour. But in 2009,
the numbers of intersections have been up to 74, in which the traffic volume of 18 intersections
is between 10,000 and 15,000 vehicles per hour. Now, the Hangkong road in Hankou district has
become the most serious traffic congested area, the traffic volume in peak hours has broken
through 20,000 vehicles per hour. In 2009, the new increased length of urban roads is 198
kilometers, which is equal to a big “parking lot”(WHTPI 2010).
Figure 3-2 Ownership of Motor Vehicles in Wuhan

(source: cited from(WHTPI 2009))
In 2008, the average traffic velocity is 23.8 k/h within the third ring route, and the average
traffic velocity of main road is 25.5 k/h (WHTPI 2009) (Figure 3-3). In 2009, within the third ring
route, the average traffic velocity is 20.4 k/h which is close to threshold of traffic jam (20 k/h).
Compared to 2008, the average traffic velocity is decrease of 2.9 kilometers per hour. By
comparing the traffic velocity of three districts in Wuhan, the fastest traffic velocity is 22.6 k/h in
Hanyang district, and the faster one is 22.2 k/h in Wuchang district. So the slowest traffic velocity
is 18 k/h in Hankou district which is of traffic jam in the whole area, and the traffic velocity of
some routes in Hankou district is lower than 15 k/h(WHTPI 2010).
Figure 3-3 Traffic Velocity in Wuhan

~ 19 ~
(source: cited from(WHTPI 2009))

Until to 2008, the total investment of transportation construction has been up to 269.2
billion RMB which accounts for 6.8% of the total GDP. Compared to 2007, the investment has
increased 29.4% in 2008. Within these investments, the total investment for constructing rail
traffic has achieved 27.5 billion RMB, which has increased 248% more than that of last year. The
investment of constructing transportation infrastructure has increased quickly, which is positive
to improve transport environment and economic development(WHTPI 2009).
3.1.2.2 Public transport development in Wuhan

In 2008, there are 20 billion public transport passengers all the year, which is an increase of
20% more than that of last year. Within these passengers, regular public bus and tram account
for 71.7% which are 14.3 billion passengers; the passenger of rail and small public bus yearly
account for 0.55% and 6.9% of the total public transport passengers respectively; the passenger
of taxi and ferry yearly account for 20.1% and 0.79% of the total public transport passengers
respectively. Therefore, the regular public bus and tram are still the major public transportation in
Wuhan(WHTPI 2009).
In recent years, the investment in public transportation is still increasing in Wuhan. In the
first phase of constructing the line 1 of rail transit, the proportion of investment is continuous
increasing. And as the construction of the first phase of line 2 and line 4, and the second phase
of line 1, the proportion of investment in public transportation is still gradually
increasing(WHTPI 2009).
3.1.2.3 Blueprint of transport development in the next 5 years in Wuhan

In Wuhan, 12 million parking spaces will be built annually in the next 5 years. The total
number of new increased parking spaces will be up to 60 million, which would basically meet the
needs of extra parking spaces. Additionally, in order to strengthen the construction of public
parking lots in central urban area, 8.5 million public parking spaces will be built in old city of
Hankou and Wuchang district, and other areas. This can gradually make up for the insufficient
parking spaces in city, and relieve the conflicts between supply and demand of parking lots in
central city area(WHTPI 2010).
With respect to the transportation access across the Yangtze River, Two-seven Bridge,
Parrot Island Bridge, the access across river of metro line 2 and line 4 will be built in the next 5
years; and some of the new bridges also will be built. All of these will form the pattern of
Yangtze River access ü “ten bridges, one tunnel, three rails” and construct the pattern of fast
road network, which aims to achieve smooth of the road in 30 minutes, and basic construction
of fast road network within major urban area(WHTPI 2010).
3.2 Characteristic of bike sharing system in Wuhan
3.2.1 Introduction of Wuhan’s bike sharing system

Wuhan launched the bike sharing system in April, 2009 which aims to satisfy people who
demand for access to “the last mile” easily. This system is operated by Xinfeida Bicycle Company
in cooperation with Wuhan Transport Planning Institute, which has been extended to 718
stations with 20000 bicycles in eight districts (expect Qingshan district). And the final objective is
to reach 2000 bicycle stations to cover the center city area in Wuhan, and the average distance of
these stations is 300 meters. Figure 3-4 shows the general information of bike sharing system in
Wuhan.
~ 20 ~
( Note: “ƽ”–yes; “ƻ”–No)

Figure 3-4 Wuhan Bikesharing System
3.2.2 Implementation of Wuhan’s bike sharing system

Since 2010, the sites selection of the new bicycle stations was operated by municipal urban
management bureau in cooperation with other institutes. In order to satisfy the different
demands of people and make the bike stations can be fully connected with other facilities in city
area, these stations are set up around bus stops, residential communities, business area,
recreational area, colleges and metro stations (Figure 3-5).
Figure 3-5 Bike rental stations in Wuhan (Wuchang)
3.2.3 Pricing of Wuhan’s bike sharing system

Because this is a convenient system for citizens, people can use this bicycle free of charge in
two hours which aims to improve the recycling use of bikes. But if people cannot return bicycle
in two hours three times, or people cannot return bicycle in 24 hours, they will be disqualified
from renting bicycles. The bicycle company supports this sharing system by advertising in the
bike stations and bikes (Figure 3-5).
3.2.4 Operations of Wuhan’s bike sharing system

This bicycle sharing system is a “manual + self-service” system (Figure 3-6) to achieve 24
hours service; users need a “public bike card” which you can get freely by using your ID card or
other valid documents. There is an intelligent box that you can pick up and return key by yourself.
~ 21 ~
Bikes can be borrowed and returned at any stations between 7:00 – 21:00. The workers of bicycle
stations in kiosks have to report to the bicycle company that the number of bikes in their stations
every night at the close time of bike stations.
Figure 3-6 Manual & Self-service system of Wuhan’s bike sharing system
3.2.5 Feedback since operations began

Since the first day of starting, the bicycle sharing system is popular with citizens and media
organizations. In a investigation, 52% people think that it is necessary to build bicycle sharing
system; 43% people think that it is important for bicycle sharing system to build two-oriented
society (resource saving society and environment friendly society); with respect to the best benefit
of bicycle system, 32% people think that is to relieve the transportation pressure in the city, 28%
people think it can save the energy consumption, 26% people think that it is convenient for short
trip. 79% people are satisfied with the service of bicycle sharing system.
But some problems are still present in this bike sharing system: the number of bicycle
stations is not sufficient, the distribution of bike stations is irrational, unbalanced capacity of
bike stations, and the equipments and bikes in bike stations are broken.
3.2.6 Future development

Bicycle Company and Wuhan Transport Planning Institute have made a plan to improve the
new added bicycle stations. First, to build newly modernized bike stations to provide public
service information; Second, to set bicycle location scientifically. Furthermore, to achieve 24
hours for borrowing and returning bikes.
~ 22 ~
4. DATA COLLECTION AND METHODOLOGY

4.1 Data collection
Based on the literature review, the research questions will be solved, and the availability of
data, road network data, land use pattern data, data on population distribution, existing bike
stations, bus stop locations, metro stations were collected. Road network provides the conditions
for evaluating the performance of bike sharing system, because the bike stations are located
along the roads and bike users can go on a trip through the road network. Concerning the land
use pattern, especially business and college area, this can be used to analyze the accessibility
between bike stations and other facilities (e.g. business, college, residential communities), based
on the idea used for planning bike sharing system. Meanwhile, because the advantage of bike
sharing system is as a feeder of public transportation, data on bus stops and metro stations will
be used to analyze the accessibility between bike stations and them. In addition, population
density is closely related to travel demand and in proportion with the amount of trip production
and attraction. With regard to these collected data, in addition to be used for evaluating the
performance of bike sharing system, these can also be used to optimize the bike sharing system
The obtained data are described as Figure 4-1.
Figure 4-1 Data description

In order to understand the characteristics of bike users, a survey should be operated. In this
study, a series of bike users’ characteristics are obtained based on questionnaires survey, see
chapter 5.1.
~ 23 ~
Figure 4-2 Road Network of width in Wuhan
Figure 4-3 Population distribution in Wuhan
~ 24 ~
Figure 4-4 Distribution of bike stations in Wuhan
Figure 4-5 Distribution of bus stops in Wuhan
~ 25 ~
Figure 4-6 Subway stations in Wuhan
Figure 4-7 Land use pattern in Wuhan
~ 26 ~
4.2 Methodology
This paper focuses on evaluating the performance of bicycle sharing system in Wuchang
district, based on the existing problems of bike sharing system in Wuhan. And method on
combination of GIS and multi criteria evaluation (MCE) is adopted to comprehensively evaluate
the overall performance of bike sharing system, by building a series of indicators which would
influence the “performance” of bike sharing system. Moreover, the bike stations with poor
performance will be optimized.
After data collection, research process were performed in according with conceptual
framework (Figure 1-3) and research design (Figure 1-5)
4.2.1 The concept of “performance” on evaluating bike sharing system

Public transit is an important component of local transportation system. Moreover,
assessing transit system performance has been one of the most widely investigated areas of
research within the transit industry(Karlaftis 2004). According to Lao et al.(2009), the majority of
existing researches focus on evaluating the performance of transit agencies from management
perspectives, little attention has been given to the spatial aspects of a public transit, such as the
characteristics of local population, transportation network, and commuting patterns.
Furthermore, Murray(2003) proposed that accessible and efficient service are vital features of
well-utilized public transit system.
In the light of these theories and researches, in this study, the “performance” evaluation is
operated in two dimensions: operational efficiency and spatial effectiveness. As follows:
x Operational efficiencyüü this will be evaluated from the perspectives of bike
users to focus on knowing the characteristics of bike users and the conditions
of bicycle lane. The objective of this dimension is to evaluate whether the
operation of bike sharing system is efficient.
x Spatial effectiveness üü this will be implemented from spatial perspective of
bike stations to focus on analyzing characteristics of bike stations and
accessibility between bike stations and other facilities. This dimension aims at
evaluating the spatial effectiveness of bike sharing system.
4.2.2 Methodology for evaluating “performance” of bike sharing system
4.2.2.1 Understanding the characteristics of bike users

The objective of understanding the characteristics of bike users is to detect operational
efficiency of bike sharing system by knowing travel purpose, travel behavior and profile of bike
users, and their attitudes to bike sharing system; these factors would be closely related to location
distribution and capacity of bike stations. In order to know of this information and data, a
survey based on questionnaires and interview with bike users and managers of bike stations was
carried out.
In this study, 24 bike stations were selected in wuchang zone of Wuchang district in field
work. In order to get a overall understanding, these stations were selected based on about 2km
space of bike stations, and different land use patterns, i.e. business area, residential community,
college, cultural entertainment area (cinema, bookstore, park, and so on), enterprise, metro
stations/bus stops. The results of field work are displayed in Section 5.1.
4.2.2.2 Methodology for the evaluation of bicycle lane

Because of the development of bicycle sharing system and the construction of bicycle
stations, the bicycle lane has been an important determinant that influences the operational
performance of bicycle sharing system. This paper focuses on evaluating bicycle lane from two
~ 27 ~
aspects: construction of bicycle lane, and operation of bicycle lane. The former focuses on
evaluating whether the network of bicycle lane is mature. And the later focuses on evaluating the
right-of-way. The more detail of analysis will be performed in section 5.2.
4.2.2.3 Methodology for analyzing characteristics of bike station

The objective of analyzing the characteristics of bike station is to deal with one aspect of
spatial effectiveness of bike sharing system. Analysis on the characteristics of bike station will be
performed by insight into spatial characteristic and non-spatial characteristic of bike stations.
With respect to spatial characteristic, it includes catchment of bike station and travel time
between bike stations. Both of catchment and travel time are closely related to the total travel
time in a trip. And non-spatial characteristic of bike station is concerned about potential
population of bike station.
z Spatial characteristic of bike station
The catchment of bike station reflects the estimated tolerance time that a person would
likely be willing to walk to a bike station. According to Krygsman et al.(2004) the catchment area
is not only a function of the absolute access and egress time but also the relative share of total
trip time. Therefore, the catchment of bike station should be calculated by using maximal
tolerance time and average walk speed, which are limited by local transportation policy and actual
traffic conditions. The results of calculation can be performed in ArcGIS
Furthermore, the travel time between bike stations reflects the interconnection of bike
stations which along the road network. And analysis of travel time is based on shortest distance
and average bike speed. Moreover, the average bike speed is defined by local transportation policy.
By using average bike speed and Network Analyst in ArcGIS, the results of travel time between
bike stations can be achieved and performed in ArcGIS.
z Potential population of bike station
Due to unavailable data on the population of using bikes, potential population is obtained
by calculating the catchment population of bike stations. And the premise of calculating
catchment population is that assume all of the inhabitants as potential population who would use
bicycles. Catchment population is calculated based on catchment area of bike station and
population distribution in study area. The result of catchment population of each bike station
can be achieved by aggregating the data of catchment area of bike station and population
distribution of study area.
4.2.2.4 Methodology for evaluating accessibility between bike stations and other facilities
Another aspect of spatial effectiveness that influences the performance of bike sharing
system is accessibility between bike stations and other facilities. Accessibility evaluation aims at
evaluating whether the spatial interaction of bike stations and other facilities (i.e. bus stops,
communities…) is effectiveness. And the accessibility is evaluated in terms of travel time.
Because the major feature of bike stations is to be integrated with bus stops, in order to
deeply detect the spatial interaction between bike station and bus stops, a method similar with
2SFCA method (Radke and Mu (2000)) is applied in this research.
x ĀBike station availabilityāüü calculating the catchment area of bike station in
maximal travel time, and searching how many bus stops around each bike station in
certain catchment area.
x ĀBus stops availabilityāüü calculating the catchment area of bus stop in maximal
travel time, and summing up the bike stations around facility’s location in certain
catchment area.
~ 28 ~
Before these two steps, the suitable travel time between bike station and bus stops should be
identified. Moreover, the theory on MSAP (maximal service area problem) will be considered to
deal with service area of facilities. The analysis can be operated by using Network Analyst in
ArcGIS, and the results would be performed in ArcGIS.
According to the principle of planning and building bike stationsü bike stations are built
around specific facilities, and 57.3% of respondents in investigation who use bikes for
commuting trips (see section 5.1.3), it is necessary to analyze the accessibility between bike
station and specific facilities (i.e. business, college, entertainment) or residential communities.
Due to there is no detailed data on the location of specific facilities, the data on land use pattern
is used to deal with this problem. The following flowchart explains the operational process of
accessibility analysis on bike stations and specific facilities.
Land use pattern

(Raster)
Selection of attribution
On specified land use pattern
Zonal Statistics
Specified land use

pattern (Raster)
Distance analysis on allocation

in Spatial Analyst
Distance of allocating
specified facility to bike
stations within 10
Figure 4-8 flowchart of analyzing accessibility between bike stations and specified facilities
4.2.3 Optimization model
4.2.3.1 Optimization of bike stations

In this study, the method of optimizing bike stations is a straightforward method based on
spatial analysis, which focuses on access and accessibility analysis. In this study, the bike stations
are seen as terminal stations and transfer stations. The optimization will be performed by the
combination of coverage model and reduction model. The objective of coverage model is to
optimize the locations of some existing bike stations, and to locate new bike stations in areas
where there are insufficient bike stations. And then, reduction model is used to reduce redundant
new bike stations. Both coverage model and reduction model are performed based on population
distribution, maximal service area, and station spacing. ArcGIS is used in the setting up of the
model.
~ 29 ~
z Coverage model
In this study, during the investigation of bike users, there are 57.3% of respondents that use
bike for commuting trips, and 55.1% and 35% of respondents using bikes integrated with public
bus and walking respectively (see section 5.1.3). Therefore, the optimization and allocation of
bike stations are completely based on the population distribution, selected bus stops which are
not covered by bike stations in certain time, and maximal service area; and make sure that every
selected bus stop and distributed population are served by bike stations, and station spacing is
lower than 30 minutes. The coverage model will be performed compliance with the following
flowchart (Figure 4-9). The more detail of coverage model will be described in section5.4.
Selecting bus stops that out
of bike stations’ coverage
Walking network Building of network dataset in

Road network
ArcCatalog
Estimation of access time

between bus stop and bike
station
Coverage of bus Network Analysis

stops (Service area)
Locating bike stations as

transfer stations
New bike
Locating bike stations as stations
terminal stations
Optimizing the location of Optimal

existing bike stations existing bike
stations
Optimal bike
stations
Figure 4-9 Flowchart of Coverage Model

z Reduction model
In order to get the optimal bike stations (including original bike stations and new added bike
stations) required for covering bus stops and population distributed in study area, a reduction
model is adopted in removing the redundant bike stations (new built bike stations). The
reduction model will be operated in accordance with the following two criterions.
x There is no bike station on express way, according to MOC(1995).
x There is no bike station as transfer station around bus stop, if there is no
population distributed in the maximal service area of bus stop.
The more detail of reduction model will be performed in section5.4.
~ 30 ~
5. ANALYSIS AND RESULTS

5.1 Characteristics of bike users
In this study, the survey was carried out based on 24 bike stations which within different
land use patterns and around different facilities, i.e. business area, residential community, college,
entertainment area (cinema, bookstore, park, and so on), enterprise, subway stations/bus stops.
And due to several reasons (e.g. limited time…), only 200 samples of questionnaires were
distributed in 24 bike stations, including weekdays and weekend. In order to know the overall
characteristics of bike users, the questions on profile of bike users and their views to bike sharing
system were included in the designed questionnaire.
5.1.1 Profile of bike users

In this survey, the profile of bike users includes users’ age, job, income, travel purpose,
rental time, the integrated travel mode, the reason for renting bikes.
According to the results of questionnaire survey, the major bike users are belong to the
major group of 20 to 29 years old (Figure 5-1), company staff ( Figure 5-2), the
income lower than 3,000 RMB (Figure 5-3). And main travel purpose is commuting trips
( Figure 5-4); majorities rent bikes between 10 and 30 minutes ( Figure
5-5). The major travel modes integrated by bike users are bus and walking (Figure 5-6). During the
survey, most of the bike users say the main reason for using bikes is convenience of bikes (Figure
5-7).
Figure 5-1 Bike Users’ age Figure 5-2 Bike Users' job
Figure 5-3 Bike Users' income Figure 5-4 Travel Purpose
~ 31 ~
Figure 5-5 Rental Time Figure 5-6 Integrated Travel Mode
Figure 5-7 the Reason for Renting Bikes
5.1.2 Bike users’ views to bike sharing system

In the questionnaire, there are five problems of bike sharing system listed:
x (ĉ) The quality of bikes is not good, and some bikes are broken;
x (Ċ) Insufficient bike stations leads to the difficulties of renting and returning bikes;
x (ċ) Irrational distribution of bike stations leads to unbalanced utilization of bikes;
x (Č) Inadequate capacity of bike stations leads to the difficulty of renting bikes in peak
hours;
x (č) There is deficient in transit information and service in bike stations;
In the light of the results, the major problems are (ĉ), (Ċ) and (Č) (see Figure 5-8). Apart
from these problems, other feedbacks were given by bike users. For instance, reducing the rental
time to improve utilization of bikes, location of bike station is not apparent, and it is difficult to
get a ‘public bike card’.
With regard to the bike sharing system in Wuhan, it is satisfied by 29.5% bike users. While
53.0% bike users say there is a lot of space to improve this system, including location and
capacity of bike stations, and managing the bike system. There are 17.5% users dissatisfied with
this system, and most of the reasons are still about bikes’ quantity and quality (see Figure 5-9).
Figure 5-8 Problems of Bike Sharing System Figure 5-9 Attitudes of users to bike sharing system
~ 32 ~
5.1.3 Conclusion
Given that questionnaire and interview with respondents and bike stations’ managers, some
basic information on bike users and bike sharing system of Wuhan is concluded:
x Major age group is between 20 and 39 years old (61%);
x Major users’ job is white-collar worker (35%);
x The income of main users is lower than 3,000RMB (73.5%);
x Major travel purpose is commuting trips (57.3%);
x Rental time is between 10 and 30 minutes (42%);
x Bike users often using bikes integrated with bus (55.1%) or walking (35.0%);
x Saving time, environmental friendly, and convenience of bikes are the major reason that
users renting bikes (67%);
x 91.5% of respondents say the bike sharing system with the problems on bikes’ quality,
and bike stations’ quantity, distribution and capacity;
x 70.5% of respondents do not satisfy the bike sharing system, and say there is a lot of
space to improve the bike sharing system of Wuhan;
x Most of the bike users responded that they wanted to find a bike station not further
than 10 minutes by walking from their home, or work place, and in 5 minutes is better.
5.2 Analysis of bicycle lane
5.2.1 Performance evaluation of bicycle lane

As discussed in section 4.2.2.2, bicycle lane is a key determinant that influences the good
operation of both bike stations and bicycle sharing system. This paper will evaluate bicycle lane
from following two aspects.
z Construction of bicycle lane network
The good operation of bike sharing system depends on the good and completed bike lane
network. This paper mainly focuses on evaluating the construction of existing bike lanes by
investigating which roads have bike lanes, to analyze whether the bike lanes network is mature.
According to MOC(1995), three-board road has two dividing strips which divide the road
into three parts. The median road is vehicle lane, and both sides of the median road are
non-motor vehicle lanes. And this rule is applied in the area with more non-motor vehicles. Due
to there is no enough data about bike lanes, this paper will identify bike lanes through finding out
three-board roads by using Google Map, and extracting these three-board roads from the data of
road network in ArcGIS. Finally, the results of data on road network with bike lanes can be
obtained (see Figure 5-10).
~ 33 ~
Figure 5-10 Bike Lane Network

In light of the results, the bike lane network is not completed and connectivity. There are
discontinuous bike lanes in blue box, which is unsafe and inconvenience for bike users to travel,
because bike users would bike on vehicle lanes and pavement if without bike lanes. Therefore, it
is necessary to build new bike lanes to make the bike lane network be improved.
z Operation of bicycle lane
The operation of bike lane is a factor that influences the performance of bike lane, which
would influence the operation of bike sharing system. The analysis on operation of bike lane will
be performed by considering the Right-of-Way of bike lane.
In Wuhan, before the operation of bike sharing system, a burning problem needs to be
solved is the Right-of-Way of bike lane, and this is the premise and basis for advocating bike
sharing system. However, the right of way of bike lane in study area has always been neglected in
the past few years. Due to the increase of motor vehicles and intensity of traffic congestion,
vehicles roads widening has been a solution to solve these problems, and this only emphasized
the right of way on motor vehicles. But the right of way on bicycles has been lost gradually.
In order to improve the Right-of-Way of bicycles, Wuhan municipal design institute has
decided to invest 55.36 million RMB to build bike lanes in 2010. Although some of the bike lanes
have been built, there are problems existing on the operation of bike lanes.
¾ There is no bike lane in new roads. So bicycles and motor vehicles have to been mixed
on roads.
¾ “Harbor bus stations” take up the space of bike lanes. A few years ago, a concept on
“harbor bus stations” was proposed, so there are some bus stations on vehicles roads
have been moved to bicycle lanes, this results in some bicycles have to be traveled on
pavement or mixed with motor vehicles on vehicle roads.
~ 34 ~
¾ Bike lanes are occupied by motor vehicle parking. Many motor vehicles park on bike
lanes; this result in people has to travel on vehicle roads and brings about the unsafe of
bicycle travel.
Bike lane
Bike lane
5.2.2 Suggestions for improving the performance of bicycle lane

In order to solve the problems of bicycle lane network and improve the actual situation of
bicycle lane, some of the suggestions are given related to the construction and operation of
bicycle lane network. As follows:
¾ The bicycle lane network should be completely constructed. Because the bicycle is
short-trip transportation (less than 6km in metropolis), the bicycle lane network should be
constructed by considering OD of trip, or the location of residential communities. This is
not only convenient for bike users traveling, but also for integrating with other
transportation systems.
¾ Management of bicycle lane should be strengthened. Because the space of bicycle lane is
taken up seriously in reality, citizens should enhance their awareness on traffic regulations,
which has a great impact on the working and operation of bicycle lane. In order to ensure
the existence of the right-of-way, the punishment for breaking traffic rules should be
strengthened.
¾ Improvement of the environment around bicycle lane. To design a beautiful environment
and travel space for bicycle lane, by combining the bicycle lane with surrounding
environment, pavement and motor vehicle lanes. This can not only provide a comfortable
space for bike users to travel, and to avoiding bad weather. In a word, to build a
comfortable and safe environment and space for bike users to travel.
5.3 Spatial analysis on bike stations
5.3.1 Characteristics of bike stations

Characteristics of bike stations play a key role in spatial effectiveness of bike sharing system.
The characteristics not only determine the performance of bike sharing system itself, but also
influence the performance of integrating with external systems (e.g. public transport system).
The elements in characteristics of bike stations can be important indicators to evaluate the
performance of bike sharing system.
5.3.1.1 Spatial characteristics of bike stations

z Catchment area of bike stations
In this study area, following questionnaire survey, 57.3% of respondents’ travel purpose is
commuting trips, and 90% of respondents use bikes integrated with walking and public bus (see
section 5.1.3), and most of the respondents responded that the access/egress time of 10 minutes
is maximal if they walk to bike stations. Apparently, bike trip has been performed as an
access/egress trip in resident trips. According to Krygsman et al.(2004) the propensity for use of
~ 35 ~
public transport deteriorates with increase in access/egress time. In the case of Wuhan, with
regards to walk to bike stations, it is assumed that people who want to walk there within 5
minutes instead of further than 10 minutes (this is derived from communicating with bike users
in my survey). In terms of the travel speed is 4km/h for walking; the threshold of travel distance
is 333m and 666m for 5minutes and 10minutes respectively.
By executing service area analysis in Network Analyst in ArcGIS, the catchment area of bike
stations within 5 minutes and 10 minutes can be obtained, see Figure 5-11. This result will be used
to calculate the potential population served by each bike station.
Figure 5-11 Catchment area of bike stations of Wuchang

z Station spacing of bike stations
Given that the survey on Wuchang district in Wuhan, 62% of respondents use bike to
transfer to public transportation, and 36% of respondents use bike as an only mode of
transportation (see Figure 5-5). Station spacing of bike stations can be determinant to
measure the efficiency of bike stations for travel, and the convenience for renting and returning
in limited time. According to MOC(1995), the standard bicycle speed for calculation is
11-14km/h, and it is necessary to use minimum value in area of traffic congestion and poor road
condition. Therefore, the bicycle speed is 11km/h in this research, due to the study area is serious
traffic congestion. MOC(1995) also proposes that the longest travel distance of bicycle should be
6km in large and medium-sized city, and 10km in small city. Based on above standard, the tolerate
travel time should be around 30 minutes in this study area.
The whole analytic process is performed in ArcGIS. First, the network dataset should be
built in ArcCatalog based on road network. Second, to operate “closest facility analysis” in
Network Analyst, bike station is defined as facility as well as incident. There is no clear network
of bike route, so the travel route is selected based on shortest route along road network. In order
to detect how many bike stations can be reached by each bike station in 30 minutes; the value of
“facility to find” should be change to 168 (total number of bike stations).In the table of “layer
properties”, the impendence is set as travel time (minutes), and default cutoff value as 30. After
the automatic running of analysis, the results of travel time between bike stations are derived.
The statistic results of analysis are shown in the following tables.
~ 36 ~
Figure 5-12 Bike stations find other stations in certain time stage
Figure 5-13 Number of bike stations without finding other stations in certain time
Based on the combination of above two pictures, 160 bike stations can find other bike
stations in 5 minutes (see Figure 5-12). Although there is a bike station without reaching other
bike stations from 5 to 10 minutes (see Figure 5-13), this bike station has already reached another
bike station within 5 minutes. So, it is obvious that all of the bike stations can reach each other
within 10 minutes (see Figure 5-13).
With respect to the bike stations (see “ID” in Figure 5-13 and Figure 5-14) which cannot
reach other bike station within 5 minutes, they can reach one or more other stations within 10
minutes (see “No. of reached bike stations” in Figure 5-14). And minimal travel time is just
around 6 minutes (see “Travel Time” in Figure 5-14).
Figure 5-14 Changes of 8 bike stations from 5 to 10 minutes
Therefore, 95% of bike stations can reach the nearest bike stations in 5 minutes; only 4.7%
of bike stations can reach the nearest bike stations around 6minutes. To sum up, the travel time
between bike stations is in accordance with the code for transport planning (which was
mentioned earlier in this section) in study area.
5.3.1.2 Catchment population of bike stations

For calculating catchment population of bike stations, data on catchment area of bike
stations (Figure 5-11) and population (Figure 4-3) has been got. However, it is necessary to
~ 37 ~
aggregate the data to get the catchment population. First, the raster data of population should be
converted to feature data—polygon, which makes the fowling activities more simple; Then,
intersection of population and catchment area, and intersected population can be obtained. After
that, summarizing the intersected population per catchment and calculating population density
per catchment. With regard to calculation of population density, first to calculate the intersected
area of intersected population per catchment, but the intersected area is just the aggregated area
of total polygon of population with intersected size instead of full size, due to intersection of
catchment area and polygon of population; then, to calculate the aggregated total area of
complete polygon of population with full size in each catchment; after these two steps, the
population density per catchment can be obtained by dividing aggregated total area by intersected
area of population. Finally, catchment population can be got by multiplying sum of intersected
population and population density per catchment (Figure 5-15). Figure 5-16 shows a general result
of potential catchment population of bike stations, it is easier to find that the number of
population served by each bike station is great different; therefore, it is necessary to distribute
bikes in according with the catchment population of bike station in reality.
Population_Wuhan
(Raster)
Conversion Raster to
Feature (polygon)
Population_Wuhan Catchment area

(polygon) of bike stations
Intersected
population
Summarizing intersected
Calculating population
population per
density per Catchment area
Catchment area
Sum of Population
intersected density per
population catchment
Multiply
Catchment
population
Figure 5-15 flowchart of calculating catchment population
~ 38 ~
Figure 5-16 Distribution of potential catchment population per bike station
5.3.2 Accessibility analysis
5.3.2.1 Bike station and bus stops

In the light of investigation on bike users of Wuhan, the maximal time of walking to a bike
station is 10 minutes (see section 5.1.3); and the main feature of bike is as an access or egress
transportation integrated with public bus. In this research, the maximum of traveling time
between bike stations and bus stops is 10 minutes, and waking speed is 4km/h. A service area
within 10 minutes can be generated in ArcGIS by using Network Analyst.
The method on analyzing accessibility between bike station and bus stops is totally in
accordance with description in section4.2.2.4. Additionally, subway stations are already contained
in the bus stops (Figure 5-17).
Figure 5-17 Bus stops containing subway stations

z Bike station availability
In this phase, the analysis is operated from the perspective of bike stations, and service
areas of bike stations are obtained by applying Network Analyst in ArcGIS. Because the
~ 39 ~
maximal walking time is 10 minutes, the search tolerance is set up as 670 meters which is equal to
multiply walking speed by 10 minutes; and settings of default breaks in “Layer Properties” is 5
and 10 minutes (see Figure 5-18). After setting up of parameters, the results of maximal service
area of bike stations can be derived (see Figure 5-19).
Figure 5-18 Settings of Layer Properties in service area analysis on bus stops
Figure 5-19 Distribution of bus stops within bike stations’ coverage
In the results, the number of bus stops covered by each bike station is lower than 7 (Figure
5-20 and Figure 5-21). Moreover, during 10 minutes, there are still 174 bus stops within service
area, and 141 bus stops out of the service area of bike stations (Table 5-1), so more bike stations
are necessary to be built. Furthermore, some of the bike stations cover 1 bus stop, while some
cover 2 to 6 bus stops, therefore, the distribution of bike stations seems sort of irrational.
~ 40 ~
Table 5-1 The results of bike station availability in 10mins

Facility Results
Service area of Bike stations 165 bike stations cover bus stops
(in 10minutes) 3 bike stations not cover bus stops
174 bus stops within service area
Bus stops
141 bus stops out of service are
Figure 5-20 Number of bus stops covered by per bike station in 10 minutes
Figure 5-21 Relationship between bike stations and bus stops (bike station avaliabity)
~ 41 ~
z Bus stop availability

Maximal service area of bus stops is also derived from network analysis in ArcGIS, which is
similar to the previous analysis on bike stations. Figure 5-22 shows the results of maximal service
area of bus stops.
Figure 5-22 Distribution of bike stations within coverage of bus stops
Table 5-2 The results of bus stop availability in 10mins

Facility Results
Service area of Bus stops 172 bus stops cover bike station
(in 10minutes) 143 bike stops don’t cover bike station
165 bike stations within service area
Bike stations
3 bike stations out of service area
With respect to the 165 bike stations which covered by bus stop in 10 minutes, there are 135
bike stations covered by bus stops within 5 minutes, and 33 bike stations covered by bus stops
during 5 and 10 minutes. Moreover, the number of bike stations covered by per bus stop is lower
than 7 (Figure 5-23), and 45% of bus stops do not cover bike stations within 10 minutes (Figure
5-24).
~ 42 ~
Figure 5-23 Number of bike station covered by per bus stop in 10mins
Figure 5-24 Relationship between bike stations and bus stops (bus stop availability)
5.3.2.2 Bike station and Residential community

Because the raster data of population is collected based on residential community, the raster
data of population is used as that of residential community for accessibility analysis. And the
process of accessibility analysis is in compliance with the flowchart (Figure 4-8) showed in section
4.2.2.4.
During the investigation of bike users, they said it is best to arrive at a bike station in 10
minutes (see section 5.1.3). Therefore, in the distance analysis, all of the facilities are allocated to
bike stations within 10 minutes.
According to the results of analysis, it is obvious that there are deficient bike stations in
study area, see red box in Figure 5-25.
~ 43 ~
Figure 5-25 Distance of allocating residential communities to bike stations within 10 minutes (in
Wuchang)
5.3.2.3 Bike stations and specified facilities

As discussed in section 4.2.2.4, the accessibility between bike stations and specific facilities
(e.g. business area, college, entertainment) should be analyzed. And the operational process of
analysis is in accordance with the flowchart (Figure 4-8) in section 4.2.2.4. As described in section
5.3.2.2 the maximal walking time between bike station and specific facility is defined as 10
minutes.
z Bike stations and business area
The data of business area is derived from raster data of land use pattern, based on
“DESCR” in attribute table. In the results of analysis, of all the business in center area of
Wuchang district are allocated to bike stations during 10 minutes, and business in red box(Figure
5-26) are not serviced by bike stations.
Figure 5-26 Distance of allocating business to bike stations in 10mins (in Wuchang)
~ 44 ~
z Bike stations and colleges

The data of colleges are derived from raster data of land use pattern based on variety of
colleges in “DESCR” of attribute table. The selected varieties of colleges are higher education
institution, secondary specialized school, and adult and continuation school.
The colleges in red box (Figure5-27) reflect that there is no adequate bike station to service
that area.
Figure5-27 Distance of allocating colleges to bike stations in 10mins (in Wuchang)

z Bike stations and entertainment
The method on extracting data of entertainment from raster data of land use pattern is
similar with college, and the selected varieties of entertainment are book stores, cinemas,
recreation, stadiums and parks. The entertainment in red box (Figure5-28) indicates that there is
an insufficient bike station.
Figure5-28 Distance of allocating entertainment to bike stations in 10mins (in Wuchang)
~ 45 ~
Figure 5-29 Distance of allocating specific facilities to bike stations in 10 minutes (in Wuchang)
z Conclusion
According to the results of accessibility analysis between bike stations and business, college
and entertainment, in general, the areas that no adequate bike stations are almost the same part in
Wuchang district (see Figure 5-29). And there are existing problems on the locations of some
bike stations, due to the spacing of two stations is too close, so it is necessary to optimize the
location of these bike stations. The more detail optimization will be shown in chapter 5.4.
5.3.3 Discussion
Based on the accessibility analysis, to some extent, it is easier to find that the
interconnection between bike stations and bus stops, residential communities, colleges is poor.
Almost half of the bus stops are not served by bike stations in 10mins (see section 5.3.2.1). And
there are deficient bike stations in some areas with high population density (see section5.3.2.2).
Moreover, by analyzing the accessibility between bike stations and specific facilities, it is easier to
find that the areas with insufficient bike stations are nearly the same part in Wuchang district (see
section5.3.2.3 ). Consequently, it is necessary to implement a optimization model to improve the
performance of this bike sharing system.
Furthermore, in this research, the data of bike stations in Wuchang district only
incorporates the bike stations which are within third ring, and which are operated by Xinfeida
Bicycle Company (main operator of Wuhan’s bike sharing system). Because the bike stations in
Qingshan zone of Wuchang district are operated by another company which only builds the bike
sharing system of Qingshan zone, then these bike stations are not included in this research.
Because the two bike sharing systems are two separate bike sharing systems, bike users cannot
rent bikes mixed. Moreover, the bike sharing system which is operated by Xinfeida Bicycle
Company is the main bike sharing system in Wuhan. Therefore, this research only focuses on the
bike sharing system operated by Xinfeida Bicycle Company.
~ 46 ~
5.4 Spatial location optimization for bike stations

As described in section4.2.3.1, the optimization will be performed in the combination of
Coverage model and Reduction mode. Coverage model is used for locating new bike stations in the area
with insufficient bike stations, and optimizing the existing bike stations. Reduction model is used
for reducing the redundant new bike stations. Moreover, the analysis process of optimization is
in accordance with the theoretical framework described in section4.2.3.1. According to the
theoretical description, the procedures of optimization as follows:
x Step 1: Selecting bus stops which are out of the coverage of bike stations, based on
section 5.3.2 Accessibility analysis.
Figure 5-30 Bus stops which do cover and not cover bike stations in 10mins
Bus stops which do not cover bike stations will be selected to be used for optimizing bike
stations.
x Step 2: Building of network dataset based on road network.
Because bike sharing system integrated with public transport system, people walk between
bus stops and bike stations. Because the walking speed is determined by physique, it is hard to
identify an accurate value of walking speed. In this study, the walking speed is defined as 4km/h
(Zuidgeest et al. 2009).
Therefore, in order to build network dataset in ArcCatalog, in the attribute table of road
network, the travel time is calculated by dividing shape length by walking speed.
x Step 3: Estimation of access time to bus stops and subway stations.
In this study, during the investigation on bike users, they said that the suitable walking time
between bike stations and bus stops is within 10 minutes. Therefore, the maximal access time is
defined as 10 minutes.
x Step 4: Network Analysis üü to calculate maximal service area of bus stops and
subway stations.
Service area analysis of Network Analyst in ArcGIS is used to calculate service area of bus
stops. Road network dataset has been created in step 2. In the service area analysis, bus stops are
defined as facilities to create service area. Because maximal access time is defined as 10 minutes,
in “Layer Properties”, the search tolerance is set as 670 meters which is equal to multiply maximal
~ 47 ~
access time by walking speed. In the previous accessibility analysis on the bus stops and existing
bike stations, 80% of bike stations were covered by bus stops within 5 minutes. Moreover, a
standard distance threshold of 0.4km was used for walking to bus stops (Zuidgeest et al. 2009).
With respect to the optimization, in order to balance the walking time between bike station and
bus stop, the “Settings of Default Breaks” in “Layer Properties” is defined as 6 and 10 minutes.
Figure 5-31 Settings of Layer Properties in Service area analysis
Figure 5-32 Service area of bus stops in 10 minutes
x Step 5: Optimizing location of some existing bike stations, and locating new bike
stations within service area of bus stops along road network, based on station spacing,
and population distribution in the study area.
In the perspective of bike stations which is divided into existing bike stations and new bike
stations, the operation of step5 will be implemented in accordance with the following parts.
x Existing bike stations —— because the locations of some bike stations are too close
which results in a bus stop is served by two or more bike stations, and some bus stops
are not served by a bike station (see red circle in Figure 5-33), this will result in irrational
utilization of bike stations for bike users. It is necessary to adjust the locations of these
bike stations slightly based on population distribution.
~ 48 ~
Figure 5-33 Problems of existing bike stations’ locations

x New bike stations üü the objective of allocating new bike stations is to deal with
the problems about the insufficient bike stations in some area. With respect to locating
new bike stations, two aspects are taken into consideration. First, to locate bike station
as transfer station in maximal service area of each bus stop. Second, to locate new bike
stations as terminal stations, if the spacing of two bikes stations beyond 30 minutes by
biking, or if there is unbalanced between bike station and population distribution in the
area.
After the analysis, the new bike stations are allocated, and existing bike stations are optimal.
However, according to the measures to locate new bike stations, lots of bike stations are located,
but some of them are redundant. In order to make the bike stations can be efficiency and
effectively utilized, it is necessary to reduce the redundant bike stations based on reduction
model.
x Step 6: Reducing redundant bike stations based on reduction model.
According to two criterions described in section4.2.2, the redundant bike stations are
removed.
The first criterion is to remove the bike stations which are along express way, because
MOC(1995) provide that bike stations cannot be built on express way. In study area, there is only
third ring of road network is express way, so the bike stations on the third ring are removed.
~ 49 ~
Third Ring
Figure 5-34 Express way of Road network

The second criterion is to remove bike stations which around bus stops, if there is no
population distribution in the maximal service area of bus stops. As figure shown (Figure 5-35),
there is no population distributed in the service area (see “blue circle” in Figure 5-35), so it is
unnecessary to build bike station around bus stop as transfer station.
Figure 5-35 Population and bus stops distributed in study area

Therefore, of all the bike stations satisfy these two criterions will be removed. After
removing redundant bike stations, the number of new additional bike stations is 174, and the
total optimal bike stations are 342. Figure 5-36 shows optimal bike stations including new bike
stations and existing bike stations. Figure 5-37 shows the distribution of bus stops and optimal
bike stations.
~ 50 ~
Figure 5-36 Optimal bike stations
Figure 5-37 Bus stops and Optimal bike stations
~ 51 ~
6. CONCLUSIONS
The objective of this research is to evaluate the performance of bike sharing system in
Wuhan in order to detect whether the bike sharing system is efficient. And based on the results
of evaluation (see section5.3.2), a method is implemented for spatial location optimization of
bike stations (see section5.4). A series of research questions have been addressed accordingly,
following the research method (see section4.2) has been proposed.
The results of performance evaluation of bike sharing system and optimized bike stations
are finally achieved following the proposed research methodology, the research questions (see
section 1.4) are answered consequently.
6.1 Achievement of this research

z Q1: what elements will be set as “characteristics” of bike uses?
Understanding the characteristics of bike users aims to detect the operational efficiency of
bike sharing system; this is closely related to the rationality and suitability of bike stations’
location and capacity. In this study, the characteristics of bike users are understood by knowing
the profile of bike users, and bike users’ views to bike sharing system. Therefore, the
“characteristics” of bike users are set as travel profile of bike users, and bike users’ view to bike
sharing system. Moreover, the profile of bike users including bike users’ age, job, income, travel
purpose, rental time, and integrated travel mode, and the reason for renting bikes. A series of
results are obtained from a survey. The more detailed results are shown in section 5.1.
z Q2: How to define and perform “characteristics” of bike stations for analysis?
The objective of analyzing the characteristics of bike station is to analyze the spatial
characteristics and non-spatial characteristic of bike stations, both of which have an effect on the
spatial effectiveness of bike sharing system. With respect to spatial characteristic, it includes
catchment of bike stations and station spacing between bike stations; and the station spacing is
performed in terms of travel time. And non-spatial characteristic of bike station is concerned
about potential catchment population of bike stations.
Furthermore, the analysis of catchment and station spacing of bike stations are operated in
ArcGIS, by using network analyst. And a computation method is implemented for calculating
catchment population of bike stations. The procedure and results are performed in section 5.3.1.
z Q3: What method will be implemented for accessibility analysis?

Accessibility of bike stations is an aspect of an aspect of spatial effectiveness that
influences the performance of bike sharing system. In this study, the accessibility analysis aims at
evaluating whether the accessibility between bike stations and other facilities (i.e. bus stops, metro
stations, residential communities, colleges, business area, and entertainment area) is effectiveness,
and the accessibility is evaluated in terms of travel time.
Accessibility analysis between bike stations and other facilities is divided into two parts; one
is for analyzing the accessibility between bike stations and bus stops or metro stations, another is
for analyzing the accessibility between bike stations and the rest facilities. And the procedure of
~ 52 ~
method on accessibility analysis was operated in ArcGIS. In light of accessibility analysis (see
section 5.3.2), it is easy to find that the accessibility between bike stations and other facilities
needs to be improved, because some of the other facilities are not covered by bike stations in
certain time.
z Q4: what criterions can be set for evaluation of bike lane?

Bicycle lane is a part of bike sharing system, and it has a great influence on the performance
of bike sharing system. In this study, the “performance” of bike sharing system is defined as
operational efficiency and spatial effectiveness, so the evaluation of bike lane is focused on the
aspects that influence the operational performance of bike sharing system. Consequently, the
criterions for the evaluation of bike lane are set as: construction of bike lane, and operation of
bike lane. The more detailed analysis is shown in section 5.2.
z Q5: Which model will be set for spatial location optimization?

In this study, the method of optimizing bike stations is a straightforward method based on
spatial analysis, which focuses on access and accessibility analysis. In this study, the bike stations
are seen as terminal stations and transfer stations. The optimization is performed by the
combination of coverage model and reduction model. The objective of coverage model is to
optimize the locations of some existing bike stations, and to locate new bike stations in areas
where there are insufficient bike stations. And then, reduction model is used to reduce redundant
new bike stations. Both coverage model and reduction mode are performed based on population
distribution, maximal service area, and station spacing. ArcGIS is used in the setting up of the
model.
6.2 Limitation of this research

The analysis on the performance evaluation of bike sharing system and optimization of
bike stations were presented gradually from data collection and analysis procedures to the analysis
results. Due to the limitation of research time and unavailable data, there are some of
deficiencies in this research, as follows:
z The survey on bike sharing system is not detailed
In this study, the survey on bike users is general, which aims to overall understand the
characteristics of bike users. Due to the limited time of investigation, the survey was mainly
operated in peak hours and focused on bike users. However, if there were enough time, the
survey can be implemented by investigating bike users evenly in peak hours and off-peak hours,
and recording the information of renting and returning bikes in survey period at the same time.
In this way, the detailed actual information of bike sharing system can be received.
z Insufficient data impair the accuracy of analysis results

In this study, only the data of road network, bus stops, metro stations, population
distribution, and land use pattern are available. The data of bike stations, colleges, community,
and business are unavailable. Due to this reason, the data of bike stations were vectorized by
combining the information of location of bike stations and Google map. And the data of
business and colleges were extracted from the raster data of land use pattern. Therefore, in the
accessibility analysis between bike stations and other facilities, the data of other facilities were not
the location of these facilities, instead of polygon data which were extracted from the raster data
of land use pattern.
~ 53 ~
Furthermore, in this research, the data of bike stations in Wuchang district only
incorporates the bike stations which are within third ring, and which are operated by Xinfeida
Bicycle Company. Because the bike stations in Qingshan zone of Wuchang district are operated
by another bicycle company which only builds the bike system of Qingshan zone, then these bike
stations are not included in this research.
To some extent, with respect to the accessibility analysis between bike stations and other
facilities, this could impair the accuracy of analysis results.
z Optimization method has an effect on the results of optimization

The optimization model was operated by combining coverage model and reduction model,
as discussed in section 4.2.3. Due to the limited research time, for locating new bike stations and
reducing redundant bike stations, a straightforward method – by manual – was proposed instead
of working out automatically by running software. Hence, in this study, the optimization model
was performed in ArcGIS based on the constraint conditions step by step, and some of the
subjective ideas were inevitable when locating new bike stations and reducing redundant bike
stations. However, adoptions of running software to operate coverage model and reduction
model automatically maybe achieve more detailed or simple results of bike stations’ distribution.
6.3 Recommendations for further research

z Quantitative traffic flow data is need to analyze the demand of bike station
Because of the unavailable data of traffic flow, the demand of bike stations was analyzed by
calculating the catchment population of bike stations and being focused on the potential bike
users. The results of analysis of catchment population could only get the general result of the
demand of bike stations. But the results cannot be seen as the actual situation of bike stations’
demand. Therefore, it is better to analyze the demand of bike stations by using traffic flow data,
which can be more practical for improving the balanced supply and demand of bike stations.
z Method of spatial location optimization of bike stations can be improved
Due to the limited research time, a straightforward method was used for optimizing spatial
location of bike stations, and it has a little limitation to generate new bike stations and reduce
redundant bike stations, and also to the results of bike stations’ location. In future research, a
spatial analysis method combined with simulation software can be used to optimize the spatial
location of bike stations, to avoid the limitation of the optimization method proposed in this
research.
~ 54 ~
LIST OF REFERENCE
Alterkawi, M. M. (2006). "A computer simulation analysis for optimizing bus stops
spacing: The case of Riyadh, Saudi Arabia." Habitat International 30(3): 500-508.
Anselin, L. (1993). Exploratory spatial data analysis and GIS. Lisbon, Portugal, West
Virginia University and University of California: 17.
Appleyard , B., Y. Zheng , et al. (2007). Smart Cities: Solutions for China’s Rapid
Urbanization: 33.
Austroads (2005). Bus-Bike Interaction within the Road Network. Sydney, Austroads
Incorporated.
Bührmann, S. (2008). Bicycles ad public-individual transport-European Developments.
Meetbike Conference. Dresden: 10.
Bührmann, S. and Rupprecht Consult Forschung & Beratung GmbH (2008). Public
Bicycles: 12.
Barclays. (2010). "BikeGrid: Cycle hire docking station viewer." from
http://www.soi.city.ac.uk/~jwo/cyclehire/stationView/.
Cervero, R. and J. Day (2008). "Suburbanization and transit-oriented development in
China." Transport Policy 15(5): 315-323.
Chien, S. I., Z. Qin, et al. (2002). Optimal bus stop locations for improving transit
accessibility. 82nd Annual Meeting of the Trans-portation Research Board. Washington, D.C.:
23.
DeMaio, P. (2008). The Bike-sharing Phenomenon - The History of Bike-sharing.
Carbusters Magazine.
DeMaio, P. (2009). "Bike-sharing:History, Impacts, Models of Provision, and Future."
Public Transportation 12: 16.
Feng , L., Y. Jiang , et al. (2009). Thinkings and Suggestions for Public Transport
Development in Chinese Cities: 14.
Fotheringham, S. and P. Rogerson (1994). Spatial Analysis and GIS, the Taylor &
Francis e-Library.
Goodchild, M. F. (2000). "GIS and Transportation: Status and Challenges."
Geoinformatica 4(2): 13.
Grotenhuis, J.-W., B. W. Wiegmans, et al. (2007). "The desired quality of integrated
multimodal travel information in public transport: Customer needs for time and effort
savings." Transport Policy 14(1): 27-38.
Hall, C. H. (2006). A Framework for Evaluation and Design of an Integrated Public
Transport System. Department of Science and Technology. Sweden, Linkopings University:
102.
ITDP-China. (2010). "China Bikesharing." from http://www.chinabikesharing.org/.
Karlaftis, M. G. (2004). "A DEA approach for evaluating the efficiency and effectiveness
of urban transit systems." European Journal of Operational Research 152(2): 354-364.
Keijer , M. J. N. and P. Rietveld (2000). "How do people get to the railway station? The
Dutch experience." Transportation Planning and Technology 23(3): 215-235.
Krygsman, S., M. Dijst, et al. (2004). "Multimodal public transport: an analysis of travel
time elements and the interconnectivity ratio." Transport Policy 11(3): 265-275.
Lao, Y. and L. Liu (2009). "Performance evaluation of bus lines with data envelopment
analysis and geographic information systems." Computers, Environment and Urban Systems
33(4): 247-255.
~ 55 ~
Li , H. and J. Guo (2010). Policy Options for Public Transport Development in Chinese
Cities. China Planning Network 2nd Urban Transportation Congress, Shanghai.
Lin, G. C. S. (2002). "The growth and structural change of Chinese cities: a contextual
and geographic analysis." Cities 19(5): 299-316.
MacIntyre, S. J. (2006). Evaluating bicycle accessibility and bike-bus integration
infrastructure : Saskatoon, SK, 2006. Geography, University of Saskatchewan Saskatoon.
Master of Arts: 103.
MAHMUD, A. R. and V. INDRIASARI (2009). "Facility Location Models Development
To Maximize Total Service Area." Theoretical and Empirical Researches in Urban
Management 4(1S): 14.
Martens, K. (2007). "Promoting bike-and-ride: The Dutch experience." Transportation
Research Part a-Policy and Practice 41(4): 326-338.
MOC (1995). Code for tansport planning on urban road. 4. Bicycle transportation. China,
Institute of Ministry of Construction of People's Republic of China.
Murray, A. T. (2001). "Strategic analysis of public transport coverage." Socio-Economic
Planning Sciences 35(3): 175-188.
Murray, A. T. (2003). "A Coverage Model for Improving Public Transit System
Accessibility and Expanding Access." Annals of Operations Research 123(1): 143-156.
Oliver. (2010, 2010-10-22). "Visualising Bike Share." from
http://oliverobrien.co.uk/2010/10/visualising-bike-share/.
PBOT. (2010). "Bike Sharing." from
http://www.portlandonline.com/transportation/index.cfm?c=50814&.
Qiu , B. (2010). Prospects of China Urban Transportation Development. China Planning
Network 2nd Urban Transportation Congress, Shanghai.
Rybarczyk, G. and C. Wu (2010). "Bicycle facility planning using GIS and multi-criteria
decision analysis." Applied Geography 30(2): 282-293.
Schroeder, B. (2010). Bicycle Sharing in China. Task Force. Kaohsiung: 20.
Tang , Y., H. Pan , et al. (2010). Bike-Renting Systems in Beijing, Shanghai and
Hangzhou and Their Impact on Travel Behavior. International Association for China Planning.
Shanghai, China, IACP.
TheBike-sharingBlog. (2009). "2009 Wrap-up, bike sharing." from
http://bike-sharing.blogspot.com/2009/12/2009-wrap-up.html.
U.S.DTFHA (1999). Guidebook on Methods to Estimate Non-Motorized Travel:
Supporting Documentation. U.S. Department of Transportation's Federal Highway
Administration 180.
UtilityCycling. (2008). "Bike Sharing Systems." from
http://www.utilitycycling.org/2010/02/bike-sharing-systems/.
Wang , Z.-g., Z. Kong , et al. (2009). "The 3rd Generation of Bike Sharing System in
Europe: Programs and Implications." Urban Transport of China 7: 6.
Warrier, N. (2010). "Summary od the process of implementing the Public Bicycle
System in Guangzhou, China." 2010, from
http://www.slideshare.net/rgadgi/guangzhou-bike-share-nitin-warrier.
Wei , X. (2010). Optimizing bus stop locations in Wuhan, China. International Institue
for Geo-information Science and Earth Observation. Enschede. MSc: 79.
WHTPI. (2009). "Wuhan Transportation Annual Report." Retrieved November,2009,
from http://www.tranbbs.com/news/cnnews/Construction/news_61895.shtml.
~ 56 ~
WHTPI. (2010). "Blue Book of Wuhan Transportation Development in 2009 "

Retrieved October, 2010, from
http://www.tranbbs.com/news/cnnews/Planning/news_74015.shtml.
WIKIPEDIA. (2010, 11 October 2010). "Bicyle sharing system." from
http://en.wikipedia.org/wiki/Bicycle_sharing_system.
XFD. (2010). "public service system of bicycle renting system in Wuhan." from
http://www.whggzxc.com/news_post.asp?ClassID=1&ArticleID=193.
Ziari, H., M. R. Keymanesh, et al. (2007). "Locating stations of public transportation
vehicles for improving transit accessibility." Transport 22(2): 6.
Zuidgeest , M., M. Brussel , et al. (2009). On bus-bike integration, SUMA.
~ 57 ~

Ying PDF

Uploaded by

Copyright:

Available Formats

Ying PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ying PDF

Uploaded by

Copyright:

Available Formats

Evaluating performance of

bicycle sharing system in Wuhan,

Thesis submitted to the Faculty of Geo-information Science and Earth

THESIS ASSESSMENT BOARD

Figure 1-1 Overview of Chinese cities' bike sharing system .................................................................. 2

Table 5-1 The results of bike station availability in 10mins .................................................................41

1.1 Background and Justification

1.1.1 Current status of transport development in China

1.1.2 Bicycle sharing system in China

Figure 1-1 Overview of Chinese cities' bike sharing system

Figure 1-2 Bicycle-Sharing System in China

1.1.3 Importance of developing a bicycle system integrated with public bus

1.2 Research Problems

1.2.1 Unbalanced supply and demand of bicycle sharing station

1.2.2 Irrational distribution of bicycle stations

1.2.3 Insufficient method for performance evaluation

1.3 Research Objectives and Questions

1.3.1 Research Objectives

1.4 Research questions

z To analyze characteristics of bike stations;

z To analyze accessibility between bike stations and other facilities;

z To select and perform indicators for the evaluation of bike lane;

Q4: what criterions can be set for evaluation?

z To establish and implement a model for spatial location optimization;

1.4.1 Conceptual Framework

Figure 1-3 Conceptual Framework

1.5 Research Design and Methods

x Understanding the characteristics of bicycle users

x Analysis of the characteristics of bike stations

x Accessibility analysis of bike stations

x Performance evaluation of bike lane

Data collection and integration Data requirement

Create database Stage 1

Analyze the coverage of bike

Analyze the catchment population Potential population

Identify indicators for accessibility Road network

Evaluating the construction and

Suggestions for improving the Bike station

Select bike station with poor Performance of

Set indicators for optimizing

Spatial location optimization

Figure 1-5 Research Design

Figure 2-1 Worldwide Distribution of Bike Sharing system

2.1.1 The development of bicycle sharing system in European Countries

2.1.1.1 History of European bike sharing system

2.1.1.2 European approaches to bike sharing system planning

Figure 2-2 Animation of bike sharing system

Figure 2-3 BikeGrid (London)

2.1.2 The development of bicycle sharing system in China

2.1.2.1 Overview of Chinese bike sharing system

Figure 2-4 Bike Sharing System in Chinese cities

Figure 2-5 Bike Sharing System in Hangzhou

2.1.2.2 The profile of Chinese bike sharing system

Figure 2-7 Pricing Model in some Chinese cities

2.2 Advantage and challenge of developing bicycle sharing system

2.2.1 Advantages of bike sharing system

2.2.2 Challenges of bike sharing system

2.3 Benefits of GIS

2.4 Service Location Model