Habitat International 82 (2018) 9–27

Contents lists available at ScienceDirect

Habitat International
journal homepage: www.elsevier.com/locate/habitatint

Urban sustainable transportation planning strategies for livable City's T

quality of life
Wann-Ming Wey∗, Jhong-You Huang
Department of Real Estate and Built Environment, National Taipei University, 151 University Road, San Shia District, New Taipei City, 23741, Taiwan, ROC

A R T I C LE I N FO A B S T R A C T

Keywords: Urban planning and its relevant transportation deploying have a particularly profound influence on the sus-
Sustainability tainability and livability of a city, and which also be crucial to the quality of life to urban residents at the same
Livability time. It was also suggested that the conception of livability should be extended to embrace the concerns asso-
Fuzzy Delphi method ciated with the sustainability. However, planning frameworks or assessment patterns that address the dynamics
Big data
of urban planning and demand for transportation deploying are relatively rare; there also few public policies in
Extendible open data
Dynamic network process
related research fields have discussed the effects of the changes in various assessment indicators over time.
Furthermore, following the rising advancements in social communication and computer technologies in modern
society, the data collection, storage, and processing capabilities of people have improved substantially. And, the
emergence of big data or extendible open data facilitates analysis and prediction availability, and enabled people
to find immediate solutions to numerous dilemmas encountered. Therefore, based on the aforementioned in-
tention, treating the city as a dynamic process with the trying of introducing the big data or extendible open data
for facilitating urban sustainability and livability is undoubtedly worth to explore in further.
The present study intends to initially examine the application of big data in sustainable and livable trans-
portation strategies in Taipei City, Taiwan. Firstly, we investigate previous research on transportation sustain-
ability in various countries to generalize our preliminary list of transportation sustainability indices that satisfy
the principles of livable cities. And, key indices were then selected through the Fuzzy Delphi Method by ad-
ministering a questionnaire to six experts from industrial, governmental, and academic sectors respectively. The
research results were applied to develop decision-making strategies for responding to the environmental dy-
namics of Taipei City's transportation infrastructure system by using the analytic network process combined with
a data-mining technique. Thus, big data pertaining to urban transportation were analyzed to predict the future
dynamic trends of the key indices and prioritize the sustainable transportation strategies for a livable city under
dynamic temporal and spatial changes. Ultimately, the policy implications of this study can not only offer a
solution for current needs related to urban planning but also serve as a more transparent decision-making or well
selection basis for developing sustainable and livable urban life in near future.

1. Introduction influential people in the world also strongly promoted a grand unified
theory which treat the cities as living organisms (Bettencourt, Lobo,
Although rapid socioeconomic development has accelerated world Helbing, Kühnert, & West, 2007; West, 2010) for understanding the
urbanization and advanced technological progress to shape the today or relationship between human engineered systems as well as the natural
even tomorrow cities (Newman, Beatley, & Boyer, 2009; Newman & environment, and attempt to guide us toward the more sustainable and
Kenworthy, 2015), it has also placed a sizeable burden and destruction livable urban life (Bettencourt & West, 2010).
on the urban sustainability and livability. Previously, in order to create Sustainability and livability are both the most important and ulti-
an urban ideal environment, Jane Jacobs, a great and extremely famous mate goals that urban life pursues as just mentioned. It was also sug-
urban scholar advocated the avant-garde concept of human-scale and gested that the conception of livability should be broadened to embrace
the beauty of urban smallness to facilitate the cities more livable and the concerns associated with the sustainability (Fidler, Olson, & Bezold,
habitable (Gans, 2006; Laurence, 2006). Recently, Professor Geoffrey 2011). To the livability and sustainability of cities, urban planning and
West who was included in Time Magazine's 2006 list of the 100 most its relevant transportation deploying have a particularly profound and

∗
Corresponding author.
E-mail addresses: [email protected] (W.-M. Wey), [email protected] (J.-Y. Huang).

https://doi.org/10.1016/j.habitatint.2018.10.002
Received 25 February 2018; Received in revised form 29 June 2018; Accepted 4 October 2018
0197-3975/ © 2018 Elsevier Ltd. All rights reserved.
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

positive effect (Taki, Maatouk, Qurnfulah, & Aljoufie, 2017; Wey, 2015; sustainable and livable urban life in near future.
Wey & Chiu, 2013; Wey, Zhang, & Chang, 2016), and which also be
crucial to the quality of life to urban residents at the same time. 2. Literature review and research background
However, planning frameworks or assessment patterns that address the
dynamics of urban planning and demand for transportation deploying 2.1. Sustainable and livable cities
are relatively rare. And, most urban planning frameworks or assess-
ments in related fields focus specifically on evaluating environmental Sustainable transportation involves applying the concept of sus-
impacts and constructing specific analytical frameworks (Deakin, tainable development to the planning and development of transporta-
Curwell, & Lombardi, 2002; Shepherd & Ortolano, 1996; Xing, Horner, tion infrastructure. For example, Black (1996) defined sustainable
El-Haram, & Bebbington, 2009) rather than considered the essence of transportation as “satisfying the current transportation and mobility
urban development as a dynamic or evolutional process. needs without sacrificing the ability of future generations to meet these
Furthermore, following the rising advancements in social commu- needs”. The OECD, Institute of Transportation in Taiwan, and Reisi
nication and computer technologies in modern society, the data col- et al. (2014) have indicated that sustainable transportation must be
lection, storage, and processing capabilities of people have improved able to respond to the concurrent and long-term environmental, social,
substantially. Additionally, the emergence of big data or extendible and economic needs and impacts. This has been supported by most
open data also facilitates analysis validity, and enabled people to find related studies. In 2009, Taiwan's Institute of Transportation further
immediate solutions to numerous dilemmas encountered. This back- explained these three dimensions of sustainable transportation. The
ground is extremely vital in utilizing the big data or extendible open environmental dimension requires considering the external effects of
data to formulate sustainable and livable transportation strategies, transportation; the social dimension requires equally satisfying the in-
which invariably change over time. Considering the time-dependency terests of residents at various socioeconomic levels when improving
of decision-making indices (Saaty, 2007) and the immediacy or ex- transportation; the economic dimension requires efficiently using and
tensibility of big data (i.e. extendible open data), we thus applied dy- conserving resources.
namic network process in conjunction with data-mining techniques to Because the abstract nature of the aforementioned principles, spe-
treat urban development as a dynamic or evolutional process (Batty, cific development strategies and assessment indices must be formulated
2017; Batty & Marshall, 2017), and the consequent changes with dy- to further assess the importance and benefits of sustainable transpor-
namic or evolutional process invariably cause shifts in demand for tation. Santos and Ribeiro (2013) indicated that indices should be
urban planning and relevant transportation deploying. In other words, created to reduce the complexity of sustainable development, maintain
the effects of such shifts over time should be considered and highlighted scientific objectivity, and improve communication processes. Studies
in order to formulate and reform more appropriate urban transporta- from various disciplines have been conducted to establish indices for
tion development strategies to our cities or even whole metropolitan sustainable transportation over the years. In 2002, the Institute of
area. Transportation in Taiwan formulated 9 goals and 47 indices encom-
Based on the aforementioned intention, treating the city as a dy- passing the dimensions of economic efficiency, environmental protec-
namic process with the trying of introducing the big data or extendible tion, and social equity according to the index categorization by the UN.
open data for facilitating urban sustainability and livability is un- Because that study did not cover all topics related to transportation, the
doubtedly worth to explore in further. In this study, we primarily intend dimension of energy use was later included as an additional dimension
to initially examine the application of big data in sustainable and li- in 2005. Subsequently, a composite sustainable transportation assess-
vable transportation strategies in Taipei City, Taiwan. This paper pro- ment index system comprising 24 quantitative and 3 qualitative indices
ceeds as follows. Firstly, we investigate previous research on trans- encompassing nine goals was formulated.
portation sustainability in various countries to generalize our Acknowledging that the selection of indices has a profound influ-
preliminary list of transportation sustainability indices that satisfy the ence on research results, Litman and Burwell (2006) aggregated con-
principles of livability. Secondly, to systematically present the opinions ventional transportation quality indices and simple sustainability in-
of all experts and integrate the consensus of all experts scientifically, dices according to their convenience and comprehensiveness and
the key indices were selected through the Fuzzy Delphi method. constructed a composite sustainable transportation index that covers
Considering the professional fields of the experts and their under- the economic, social, and environmental dimensions of sustainable
standing of the transportation status of Taipei City, two experts from transportation. To identify a framework that integrates the information
the industrial, governmental, and academic sectors were selected (six in from multiple sources to evaluate transportation sustainability, Awasthi
total) to answer the questionnaires. Thirdly, the existing temporal and Chauhan (2011) incorporated the analytic hierarchy process to
quantitative data of the seven key sustainability and livability indices define nine indices that involved the social, transportation, environ-
which constructed by Fuzzy Delphi method were collected. Then, the mental, energy, and economical dimensions of sustainable transporta-
time-series analysis was performed to establish an autoregressive in- tion; subsequently, numerous data sources (e.g., expert questionnaires,
tegrated moving average model (ARIMA) for each index in order to sensors, and models) were applied to evaluate sustainable transporta-
calculate the rate of change in each index from the 2009/Q2 to the tion strategies. To assess sustainable transportation strategies im-
2016/Q4 (31 quarters in total). The rate of change in each index would plemented in Taipei City, Shiau (2012) proposed a composite index that
be applied to reveal the dynamic weight changes in data-oriented dy- incorporated the social, economic, environmental, energy, and financial
namic network process to express the dynamic concept of indices or dimensions of sustainable transportation.
alternatives that change over time. Thus, big data pertaining to urban After integrating the transportation indices from previous studies,
transportation were analyzed to predict the future dynamic trends of Santos and Ribeiro (2013) selected 20 indices for the social, economic,
the key indices and prioritize the sustainable and livable strategies for and environmental dimensions of sustainable transportation and pre-
our empirical area under dynamic temporal and spatial changes. On the liminarily determined their importance on the basis of how frequently
other hand, the research results were applied to develop decision- they appeared in previous studies. For example, CO2 emissions, energy
making strategies for responding to the environmental dynamics of consumption, traffic accidents, safety, and accessibility were key in-
Taipei City's transportation infrastructure system by using a dynamic dices that had appeared in approximately 70% of the investigated
network process combined with a data-mining technique. Ultimately, studies. Finally, cases of urban development were used to examine the
the policy implications of this study can not only offer a solution for applicability of sustainable development indices in Rio de Janeiro. To
current needs related to urban planning but also serve as a more eliminate unnecessary indices, Reisi, Aye, Rajabifard, and Ngo (2014)
transparent decision-making or well selection basis for developing reviewed previous studies and estimated the existing data of

10
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

Melbourne, Australia, to create nine transportation sustainability in- knowledge discovery activities (Storey & Song, 2017). For example, in
dices encompassing environmental, social, and economic dimensions; the context of smart sustainable cities, big data analytics denotes a
these indices were then integrated into a single composite index. collection of sophisticated and dedicated software applications and
In addition, regarding the urban livability, Jane Jacobs probably database systems run by machines with very high processing power,
was the most famous proponent and influential activist in the history of which can turn a large amount of urban data into useful knowledge for
American city planning. In The Death and Life of Great American Cities, well-informed decision-making and enhanced insights pertaining to
published in 1961, Jacobs advocated the construction of livable cities various urban domains, such as transport, mobility, traffic, environ-
with abundant street life and commercial and cultural diversity. She ment, energy, land use, planning, and design (Bibri & Krogstie, 2017).
stated that only a diverse human-scale community can lead to the Despite attention and embrace from many scientific researchers or
foundation of a simple yet great city. Subsequently, many government fields, the definition of big data is still highly controversial and remains
agencies or research institutions in recent also put forward their own ambiguous. One of the major widely debate is that the meaning of big
understanding and views. For example, in Cities in a Globalizing World: data is not accurate enough. Due to big data which may means data
Global Report on Human Settlements, the first global report of the United itself are too big or large that could not be processed easily in the past,
Nations (UN) Human Settlements Program stated that the provision of but which maybe could calculate smoothly through the amateur soft-
public infrastructure, equal opportunities for medical and education ware and hardware now (Manovich, 2011). Thus, big data itself is, in
services, and safe and comfortable community environments are the many ways, a poor term with no precisely defined (Boyd & Crawford,
prerequisites for building a livable city. The U.S. Metropolitan Planning 2012; Manovich, 2011). On the other hand, the meaning or definition
Commission (MPC) reported that a city must develop safe, reliable, and of big data suggests that data are too big or large only in relation to our
economically feasible transportation options to lower the costs of current computational power rather than exact defining characteristic
household transportation, reduce national reliance on imported raw of this new data ecosystem (Floridi, 2012). In fact, some of the big data
petroleum, and improve air quality. In Singapore, the Urban Land that we praise highly now (e.g., Twitter messages about a particular
Institute and Center for Liveable Cities also maintained that en- topic) are not even larger than the census data (e.g., government open
vironmentally friendly transportation systems must be established to data about the socio-economic survey) that were not recognized as big
eliminate traffic pollution and congestion, which undermine the quality data in the early days (Boyd & Crawford, 2012). As described by Pro-
of a city's livability. fessor Michael Batty, a renowned British urban planner and scientist of
Moreover, Evans (2002) addressed livelihood and ecological sus- complex urban systems, big data are data bodies that expand con-
tainability as two aspects of city livability. Livelihood indicates sa- tinually and thus continue to challenge the capacities of data analysis
tisfactory living conditions. When the lifestyles of city residents cause tools (Batty, 2013).
environmental decline, the quality of life of residents is diminished. A According to the discussion aforementioned, we thus find another
livable city satisfies the livelihood needs of all its residents under the major widely debate is that big data should not only refer to very large
prerequisites of ecological environmental protection. Fidler et al. data sets, but also to the tools and procedures used to manipulate and
(2011) also emphasized that many urban illness or social threats, in- analyze them (Boyd & Crawford, 2012). In other words, the most im-
cluding urban sprawl, rising energy costs, climate change, and fiscal portant value of big data does not rely on the amount or quantities of
challenges are likely to work against efforts to make communities more data itself. As described by Professor Geoffrey West, a renowned British
livable. On the basis of these findings, we should develop strategies that theoretical physicist and scientist of grand unified theory of sustain-
counter these negative illness or threats while also improving our urban ability (West, 2010), science is meritocratic and not all data are equal
life. In the most recent period, Zhan et al. (2018) conducted a large- (West, 2017). In some extreme situations, we even must try to eliminate
scale questionnaire survey on 40 major cities in China in order to fill the large part of data sets that is useless and may confuse experimental
gaps of past studies for urban livability. The results show that the six results (e.g., the discovery of the Higgs boson and top quark). Thus,
dimensions of urban livability (including public facilities, natural en- maybe for most scientific fields, we have to put our main effort and
vironment, socio-cultural environment, urban security, environmental focus on exploring or mining critical parts that hidden in data sets via
health, and convenient transportation) have significant and positive design techniques, analytical tools, execution procedures, decision
impacts upon overall satisfaction with urban livability, of which the support systems, or even creative solution algorithms (Chang, Wey, &
natural environment, convenient transportation, environmental health Tseng, 2009; Janssen & Kuk, 2016; Lai & Huang, 2017; Wey & Wu,
are the greatest contributing factors. 2007) rather than the dominate characteristic of 5Vs (volume, velocity,
In summary, most studies have maintained that a livable city must variety, veracity, and value) or precise definition of the big data. Then,
encompass numerous aspects such as economic development, social it thus would inform or even enhanced decision making as well as fu-
security, environmental conservation, history and culture, and trans- ture strategies we needed, especially in the situation of new urban data
portation. And, transportation planning is crucial for constructing streams generating rapidly (Batty, 2012).
public infrastructure; the diversity, safety, convenience, and sustain- Another major widely debate about big data or large dataset is that
ability of transportation activities are vital for a strong, fair, and effi- may lead to decision paralysis, or at least decision delay, that is, the
cient living environment as well as the livability of a city. On the other allure of data encourages decision makers to wait more possible in-
hand, sustainable transportation is any transportation mode that sa- formation (Håkonsson & Carroll, 2016). And, the delayed decisions may
tisfies current transportation needs without sacrificing the needs of ultimately not be applicable to rapidly changing urban systems (Batty,
future generations, the evaluation of which involves environmental, 2012). Although it seems that big data makes it possible for us to collect
social, and economic dimensions. The environmental dimension em- more data to find more useful information, the truth is that more data
phasizes ecological conservation and alternative energy development; do not necessarily mean more useful information. It may contain more
the social dimension emphasizes equality and safety; the economic di- ambiguous or abnormal data (Tsai, Lai, Chao, & Vasilakos, 2015; West,
mension emphasizes the balance between cost and efficiency and the 2017). Therefore, bigger data (more data) that we wait for or search for
improvement of economic competitiveness. are not always better data (Boyd & Crawford, 2012). Moreover, the
above decision scenarios also specifically alert that there is a tendency
2.2. Big data and its applications pursuing more data that just due to decision makers could (easily)
obtain, not because the data would facilitate or enable decision makers
The popularity and tide of big data have resulted in the develop- to make the right or better decisions (Håkonsson & Carroll, 2016).
ment and applications of technologies and methods aimed at effectively As discussed above, although there are still many questions and
using massive amounts of data to support decision-making and controversies that still need to be resolved, the cooperation between big

11
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

Fig. 1. The goal and research design of this study.

data and urban development is still booming and continues to advance. developments in Information and Communication Technology in fur-
Numerous types of datasets are applied to research on urban and ther, and logistics will play a critical role in future cities due to the
transportation planning. Considering the importance of dynamic in- increasingly micro-dynamic nature of socio-economic and globalised
formation in understanding urban spaces, Soto and Frias-Martinez production and consumption patterns. Schlingensiepen, Mehmood, and
(2011) used mobile phone data from urban spaces and defined five Nemtanu (2015) and Schlingensiepen, Nemtanu, Mehmood, and
clusters of land use. Examples of big data being applied to research on McCluskey (2016) also emphasized the application and benefits of
travel routes include time-series global positioning system data being cutting-edge technology. Coincidentally, they both believed that future
used to analyze the travel routes of taxicabs (Zheng, Liu, Yuan, & Xie, cities will be driven through advanced complex ICT systems (i.e., au-
2011), smart card systems—the usage of which has increased as a fee- tonomic transport management system) that harness the power of data
payment mechanism in urban public transportation—have been em- collected from distributed network of sensors in order to provide per-
ployed in several studies to investigate the temporospatial dynamics of sonalized information and services to its consumers under sustainability
urban public transportation trips (Bagchi & White, 2005; Pelletier, and other constraints. In similar way, Naim and Rashid (2015) declared
Trépanier, & Morency, 2011). Social media data, although having enterprise systems and supply chain management systems as a critical
emerged only recently, have expanded rapidly and now provide a part of smart city setting could facilitate sustainability and have a po-
massive source of big data for real-time information on travel patterns, sitive impact on corporate financial returns. Therefore, enterprises
such as transportation information from the keywords in social media should pay more attention to the three dimensions (i.e., economy, so-
(Gal-Tzur et al., 2014). ciety and environment) of sustainability.
Recently, it is worth to note that various scientific innovations in Due to the industry-driven, top-down, design approaches com-
urban development field are flourishing especially between cutting- pletely ignore the self-organization, the self-sustaining structures, and
edge technologies and its emerging requirements (Alazawi, Alani, the complexity of urban systems (Batty, 2008; Batty & Marshall, 2017;
Abdljabar, Altowaijri, & Mehmood, 2014; Mehmood & Graham, 2015; Bettencourt et al., 2007), Naim and Rashid (2015) also deeply reflected
Suma, Mehmood, Albugami, Katib, & Albeshri, 2017). Often of these that the ICT-based and industry-driven approaches maybe are not the
cutting-edge technologies and its emerging requirements claim that only solution for urbanization and future city designs. Based on the
they could promote the sustainability and livability of the city or even above insightful rethinking and reflecting, we thus realize that the
make the city, society or environment smarter. For example, Ahmad common ground and means of ICT-based or industry-driven approaches
and Mehmood (2016) suggested that future cities are driven by the often tend to directly achieved sustainability or livability through

12
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

innovative technologies, mechanisms or intelligent systems. When en- (a) Taipei City is the political, economic, and cultural center of Taiwan.
joying great technological convenience and business benefits afore- The total population of Taipei City in October 2014 was 2,697,644.
mentioned, however, we have to notice that the implementation of Data from that year show that residents owned 286 cars and 385
these innovative technologies, mechanisms or even intelligent systems motorcycles per 1000 population in 2013. Thus, the number of
requires a lot of data resources, hardware architecture, construction private vehicle trips in Taipei City is high. Moreover, the population
cost as well as probably facing more unknown risks or complexities. concentration is also high, and residents experience difficulty
Comparing to traditional planning, these innovative technologies, me- commuting between New Taipei City and Taipei City. These
chanisms or intelligent systems may be necessary to assessment care- transportation needs require immediate attention.
fully via relevant experts or research institutions in further. (b) Urban development strategies for Taipei City must promote har-
mony between humans and nature, a balance between the cityscape
and nature, and macroeconomic objectives, social harmony, and
3. Research design
ecologically virtuous cycle. Therefore, the Taipei City Government's
vision for future development is focused on foresight, pleasure, li-
The goal of this study was to establish transportation sustainability
vability, culture, ecology, information, and safety. White Paper on
indices for a livable city, to predict their future dynamics, and to for-
Transportation Policies from the 2020 Program for the Long-Term
mulate sustainable urban transportation strategies that fulfill the time-
Development in Taipei lists sustainable transportation, affinity and
dependency of the indices. The research design (Fig. 1) is described in
humanism, and livability and safety as the visions of development
the following four steps:
of the city government for formulating the development goals and
Step 1 Collect and review literature on livable cities and sustainable
executing the policy strategies.
transportation worldwide. Examine and analyze the literature to gen-
(c) On the actual transportation construction, major development plans
eralize a list of transportation sustainability assessment indices for a
for the city's public transportation network have neared comple-
livable city.
tion. In 2013 and 2014, the Xinzhuang, Xinyi, and Songshan lines of
Step 2 Select the applicable indices for the proposed research model
Taipei's mass rapid transit (MRT) network were opened, increasing
and the conditions of the area of empirical study through the Fuzzy
the total length of the MRT to 129.2 km. In 2014, daily peak traffic
Delphi Method (FDM). Generalize the dependency relations among the
was reached 1.63 million trips. The Taipei City Government has
indices for the subsequent modeling process.
promoted a dynamic bus information system to improve the quality
Step 3 Use data mining to analyze the transportation data to predict
of bus services by fitting buses with GPS tracking systems and
dynamic changes in the indices over time. Operationalize these changes
communication devices to provide real-time information on bus
in the DNP by examining the priorities of the urban transportation
travel times for seamless information services.
sustainability indices to develop a questionnaire survey for a panel of
experts. Analyze the results of the data-oriented and expert-oriented
The Taipei City Government has also promoted the Taipei Bike
DNPs and compare the findings. For further application of the rates of
Sharing System Establishment and Management Program to encourage
changes in the index data and to identify current index weights and
people to use bicycles for short trips. A network of urban bicycle lanes
predict future weight changes, this study investigated the raw data
has also been integrated into the bike sharing system to encourage
through a time-series analysis, a data-mining tool that is frequently
people to use low-pollution, low-energy public bicycles instead of pri-
used to predict continuous variable processes. This was employed to
vate motor vehicles for short trips, thereby mitigating traffic conges-
establish a prediction model for each index. The DNP proposed by Saaty
tion, environmental pollution, and energy exhaustion.
(2007) can be applied in sturctural or functional form. The structural
In summary, the current transportation policies implemented by the
form incorporates different time points as alternative programs and
Taipei City Government are oriented toward sustainable transportation,
compares the criteria with the programs in pairs. The functional DNP
and the orientation of the urban development also satisfies the visions
involves using the standard dynamic judgment matrix A(t), which ex-
of a livable city. Thus, Taipei City is within the scope of this study.
plicitly includes the functions of time. The typical form of a judgment
Additionally, the city's transportation demand is high, and its fully
matrix in dynamic form is:
developed transportation infrastructure provides a massive source of
transportation data for follow-up research procedures. Therefore,
⎡ a11 (t ) ⎤ a1n (t ) 0, =.
Taipei City was selected as the area of the empirical study to investigate
⎢ a (t) ⎥…… a2n (t ) 0, =.
A(t) = ⎢ 21 ⎥… the key indices and strategies satisfying Taipei City's livability and
⎢ ⋮ ⎥… ⋮ 0, =.
transportation sustainability.
⎣ an1 (t)
⎢ ⎦ ann (t )
⎥ 0, =.

As in the discrete case, when A(t) is consistent, we have

4.2. Establishing the target system and preliminarily selecting the indices
aij (t ) = wi (t )/ wj (t ) . The functional DNP, which explicitly incorporates
time variables, clarifies the dynamic weight curve of each factor ac-
The following figure shows the sustainability indices drawn from
cording to the relative rate of change in each factor over time, the
the discussed studies and applied to the livable city concept (Fig. 3) for
findings of which are then assessed by experts.
the follow-up FDM and empirical analysis. The preliminarily selected
Step 4 Generalize and discuss the processes and results of the em-
indices were based on the environmental, economic, and social di-
pirical analysis and propose suggestions on formulating sustainable
mensions. The leading indices were CO2 (greenhouse gas) emissions per
transportation strategies for the livable city examined in this study.
unit, nonrenewable energy and fuel consumption, and air and noise
pollution for the environmental dimension; infrastructure, congestion
4. Empirical study and analysis severity, and household expenses on transportation for the environ-
mental dimension; and accessibility to public transportation and safety
4.1. Area and range of the empirical study for social dimension. Other indices representative of transportation
sustainability were the density of land use, transportation diversity,
An area that fulfills the goal and model of this study was selected for land consumption in transportation, and the use of public and non-
the empirical study. Considering the environmental, social, and eco- motor vehicles. Table 1 lists a description of the indices.
nomic dimensions under investigation, Taipei City (Fig. 2) was selected
as the area of the empirical study for the following reasons:

13
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

Fig. 2. Administrative districts of Taipei City.

Fig. 3. Research framework of a livable city concept.

14
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

Table 1
Descriptions of the transportation sustainability indices for a livable city.
Dimension Index Description

A.Environment A1. Greenhouse Gas Emissions Effective control and monitoring of greenhouse gas emission in urban transportation to reduce damage to the
ecological environment and human health, thereby achieving the sustainable development of urban living and
the natural environment.
A2. Land Consumption in Transportation Effectively plan and manage land use in transportation to improve land resource use efficiency and mitigate the
environmental impact of transportation, thereby safeguarding the quality of daily living and the natural
environment.
A3. Energy Use Develop renewable energy to reduce people's reliance on nonrenewable energies (e.g., fossil fuels) to enabling
effective energy recycling, mitigate the impact of nonrenewable energy use on the environment and human
health, and attain the goals of livability and sustainability.
A4. Air and Noise Pollution Effectively control and monitor noise and emissions from urban traffic to mitigate damage to the ecological
environment and human health, thereby achieving sustainability in both the urban environment and the natural
environment.
A5. Public and Non-motor Vehicle Use Promote the use of public and non-motor transportation vehicles and reduce the use of motor vehicles to
mitigate the impact of vehicles on the natural environment and attain the sustainable development of urban
daily living and the natural environment.
A6. Density of Land Use Promote high-density land use to reduce travel distances and times to lower unnecessary resource use and
transportation trips and thereby improve the urban environment.
B.Economy B1. Infrastructures Promote the transportation quality, convenience, and livability of the city through constructing a
comprehensive range of transportation infrastructure services to achieve the objectives of economic efficiency
and sustainability in urban transportation.
B2. Congestion Severity Provide diverse transportation vehicles, encourage minimal use of private vehicles, and strengthen the overall
transportation system network to reduce traffic congestion and improve the livability of the city, and thereby
achieve the objectives of economic efficiency and sustainability in urban transportation.
B3. Household Expenses on Transportation Develop diverse, safe, reliable, and economically feasible transportation options to reduce household expenses
on transportation and promote the quality of daily living, thereby improving the livability and transportation
efficiency of the city.
B4. Government Investment on Improve the efficiency of the government's investment in transportation infrastructure and policy, lower the
Transportation cost per unit of the investment, and prioritize investment in environmentally friendly transportation systems
and infrastructure to improve the efficiency of resource use.
B5. Mixed and Use Promote mixed land use to shorten travel distances and reduce the demand for commuting to reduce household
expenses on transportation and improve transportation efficiency.
B6. Technology Application and Management Upgrade technology, improve the energy use and capacity of transportation units, and enhance the efficiency of
transportation system operations, monitoring, information management, and law enforcement through
innovative governance (e.g., promote the use of hybrid vehicles and smart transportation systems).
C.Society C1. Diversity of the Transportation System Provide various types of transportation vehicles to satisfy the transportation needs of different groups to achieve
social sustainability and enhance the livability of the city.
C2. Quality of the Transportation for Consider and satisfy the basic transportation needs of disadvantaged groups (e.g., persons with disabilities,
Disadvantaged Groups children, elderly citizens, low-income earners, and the residents in remote areas) in designing transportation
systems to promote the livability of the city for all people and maintain the equality in urban development.
C3. Accessibility of the Public Transportation Improve the public transportation network and its accessibility and convenience to reduce the use of private
vehicles, lower traffic congestion, and satisfy the basic transportation needs of all people, thereby attaining the
livability of the city and the equality in its development.
C4. Safety Enhance road and sidewalk safety to minimize the risks of traffic accidents and casualties to mitigate the
consequent social damage, thereby achieving social sustainability and enhancing the livability of the city.
C5. Health Impact Reduce traffic pollution and accidents to mitigate their damage to public health, thereby improving the
livability of the urban environment and sustainable transportation development.
C6. Satisfaction Improve the of transportation infrastructure (e.g., roads and public and non-motor transportation vehicles) to
raise public satisfaction, thereby enhancing the livability and sustainability of the urban environment.

4.3. Using the FDM to select the transportation sustainability indices that transportation (Index C3) versus use of public and non-motor vehicles
fulfill the goals of developing Taipei City as a livable city (Index A5), with a difference of 0.34; (b) health impact (Index C5)
versus diversity of transportation systems (Index C1), with a difference
To establish the transportation indices and strategies and objec- of 0.19; (c) mixed land use (Index B5) versus technology application
tively fulfill the demands of the area of the empirical study and the and management (Index B6), with a difference of 0.71. To obtain a
considerations of planners to identify the most appropriate and efficient suitable number of selected indices for the follow-up research proce-
indices and strategies, the transportation sustainability indices dures (total number of indices selected = 7), Index C5 (health impact)
(Table 1.) were examined through an FDM expert questionnaire to af- was selected as the threshold in this study (threshold Value = 7.30).
firm the objectivity of the indices for further strategic planning. Con-
sidering the professional fields of the experts and their understanding of
the transportation status of Taipei City, two experts from the industrial, 4.4. Predicting the dynamic trends of the transportation sustainability
governmental, and academic sectors were selected (six in total) to an- indices through data mining
swer the questionnaires.
Table 2 depicts the questionnaire results. The expert consensus Temporal quantitative data of the seven selected transportation
value Gi is > 6 for most of the indices. In other words, all 18 indices sustainability indices were collected. A time-series analysis was per-
were considerably representative of transportation sustainability in formed to establish an autoregressive integrated moving average model
Taipei City. However, for the operability of future studies, these indices (ARIMA) for each index in order to calculate the rate of change in each
were selected from the steepest region of the consensus curve as the index from the 2009/Q2 to the 2016/Q4 (31 quarters in total). Because
threshold value. The three regions with the largest differences in Gi of the limitations in acquiring the index data, only the actual available
values on the curve are listed as follows: (a) accessibility of public data could be used as the index application data.
For the operation and prediction consistency, the unit of

15
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

Table 2
Index assessment results in the FDM expert questionnaire.
Index Conservative Optimistic Singular Value (a) Geometric Means Certified Expert Consensus Value
Value (Ci) Value (Oi) Value (Zi) (Gi)

min max min max min max Ci Oi Singular

Value

A1. Greenhouse Gas Emissions 4 7 6 9 5 8 5.67 8.30 6.90 1.63 6.99

A2. Land Consumption in Transportation 6 8 8 10 7 9 7.13 9.47 8.14 2.33 8.30
A3. Energy Use 5 6 8 9 7 7 5.79 8.19 7.00 4.41 6.99
A4. Air and Noise Pollution 4 8 6 10 5 9 5.71 8.19 6.53 0.49 6.95
A5. Use of Public and Non-motor Vehicles 4 9 8 10 6 9 6.24 8.94 7.64 1.71 7.59
A6. Density of Land Use 4 8 6 10 5 9 5.68 8.21 6.88 0.53 6.95
B1. Infrastructure 4 7 6 10 5 8 5.67 8.28 6.90 1.60 6.98
B2. Congestion Severity 4 8 8 10 6 9 5.96 8.79 7.39 2.83 7.37
B3. Household Expenses on Transportation 4 8 6 10 5 9 5.60 8.07 6.99 0.47 6.83
B4. Government Investment on Transportation 4 8 8 10 6 9 6.19 8.81 7.44 2.61 7.50
B5. Mixed Land Use 4 6 6 9 5 7 5.28 7.93 6.29 2.65 6.60
B6. Technology Application and Management 4 7 6 9 5 7 4.68 7.11 5.75 1.43 5.89
C1. Diversity of Transportation Systems 4 8 6 10 5 8 5.99 8.23 6.90 0.24 7.11
C2. Quality of Transportation for Disadvantaged 4 8 6 10 5 9 5.92 8.23 7.31 0.31 7.08
Groups
C3. Accessibility of Public Transportation 4 10 8 10 6 10 6.74 9.12 8.07 0.38 7.93
C4. Safety 5 8 9 10 7 9 6.74 9.65 8.30 3.91 8.20
C5. Health Impact 4 8 7 10 6 9 6.17 8.43 7.42 1.26 7.30
C6. Satisfaction 4 8 6 10 5 9 5.79 8.38 7.03 0.59 7.08

Note: Grey-highlighted areas indicate the indices that fulfilled the threshold.

Table 3
Descriptions of the actual application data of the indices.
Index Measurement Data Relationship with the Index Data Source

A1.Land Consumption in Ratio of the amount of land used for Reveals the extent of the importance that the city － Taipei Statistical Database
Transportation transportation to the total amount of land government associates with transportation － Department of Urban Development, Taipei
for public infrastructures construction and the traffic status possibly seen in the City Government
city.
A2.Use of Public and Non- Rate of public and non-motor The promotion and use of public and non-motor － Statistics Inquiry, Ministry of
motor Vehicles transportation vehicle use transportation vehicles are key criteria for the Transportation and Communications
livability of a city. － Taiwan Railways Administration, Ministry
of Transportation and Communications
－ Taipei Rapid Transit Corporation
－ Taipei Statistical Database
－ Taipei Transportation Statistical System
B1.Congestion Severity Number of cars/motorcycles per 1000 The total number of cars and motorcycles reflects the － Taipei Statistical Database
population severity of traffic congestion at a certain extent. － Taipei Transportation Statistical System
B2.Government Investment Budgets of the Taipei Department of The efficiency of government investment on － Taipei Statistical Database
on Transportation Transportation transportation construction directly influences the － Department of Urban Development, Taipei
quality of urban transportation. City Government
C1.Accessibility of the Public Numbers of MRT stations, joint bus The number of public transportation stations directly － Taipei Transportation Statistical System
Transportation routes, and YouBike stations affects the intentions of people to use public
transportation.
C2.Safety Average monthly number of traffic Reducing the number of traffic accidents promotes － Taipei Transportation Statistical System
accidents urban transportation safety to at a certain extent. － Traffic Division, Taipei City Police
Department
C3.Health Impact Standard index average monthly Air quality is crucial to the health and quality of life of － Environmental Quality Data Storage
pollution levels people. Transportation waste emissions are a major System, Environmental Protection
cause of urban air pollution. Administration

observation was set as one month; data observations from January 2009 represents the lag; ai is the coefficient of yt − i ; εt represents white noise.
to December 2014 (60 data observations in total) were prioritized. q
Table 3 illustrates the data on the actual application of the indices, how yt = a0 + ∑ bi εt−i
the data are related to the indices, and the data source. i=1 (4.2)
The ARIMA combines two types of data-generating process models,
In Formula (4.2), a 0 represents the intercept of the constant; p re-
namely the autoregressive (AR) model and the moving average (MA)
presents the lag; bi is the coefficient of εt − i ; εt represents white noise.
model. Formula (4.1) shows the generalized AR (p) model; Formula (4.2)
depicts the generalized MA (q) model; and Formula (4.3) displays the p q

formal definition of ARIMA (p, q) (Yang, 2010). yt = a0 + ∑ ai yt−i + εt + ∑ bi εt−i

i=1 i=1 (4.3)
p
Each index can be used to generate multiple models with a good fit.
yt = a0 + ∑ ai yt−i + εt
i=1 (4.1) The model with the closest fit for each index was determined according
to the Akaike information criterion (AIC) and the Schwarz Bayesian
In Formula (4.1), a 0 represents the intercept of the constant; p criterion (SBC). The smaller the AIC and SBC were, the higher the

16
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

Table 4 4.5.1. Verifying the data-oriented DNP

Best-fit model and formula derivation of each index. The data-oriented DNP was divided into three steps. First, the rates
Index Index Application Data Finalized of changes in the indices were compared in pairs to determine the
ARIMA (p, d, q) corresponding level of importance of each index in achieving the goal of
city livability. Second, the indices with internal dependency were
A1 Ratio of the amount of land used for ARIMA (1,1,1) compared in pairs to determine the extent of internal dependency.
transportation to the amount of land used for all
Third, the assessment matrix derived from the paired comparison was
public infrastructure
A2 Use of public and non-motor vehicles ARIMA (4,1,5) multiplied by the weighted vector of each index to calculate the weight
B1 The number of cars and motorcycles per 1000 ARIMA (0,1,2) of each index under the goal of city livability. These steps were repeated
population to obtain the weights of each index at different time points (t).
B2 Budgets by the Taipei Department of ARIMA (0,1,1)
Transportation
C1 The number of MRT stations, bus routes, and ARIMA (1,1,1)
4.5.1.1. Weights of the indices without considering their internal
YouBike stations dependency (w1). The quarterly rates of change in each index in the
C2 Rate of vehicle accidents (cases/10,000 ARIMA (3,1,(1,3)) same year were compared in pairs to obtain their weights without
vehicles) considering their internal dependency; the changes are represented as
C3 Pollutant standards index ARIMA (3,1,(5))
vector w1. The formula of the paired comparison is expressed as follows:

Δx
goodness-of-fit of a model was considered to be. When the AIC differed
w1xy =
Δy (4.8)
considerably from SBC, the models with the closest fit were selected
according to their parsimony. The AIC and SBC were calculated as where w1xyrepresents the weight of the index X relative to index Y
shown in Formulas (4.4) and (4.5). (without considering internal dependency); Δx and Δy respectively
represent the amounts of changes in X and Y. For a simple
AIC = T ln(SSE ) + 2k (4.4)
comparison of the rates of changes in the two indices, Δx was divided
SBC = T ln(SSE ) + k ln(T ) by Δy , and the absolute value was obtained.
(4.5)

where ln(SSE ) is the natural logarithm of the sum of squared errors; 4.5.1.2. Index internal dependency weighting matrix (W2 ). In the data-
ln(T ) is the natural logarithm of the total number of samples; k re- oriented DNP, the quarterly rates of change in each index in the same
presents the total number of parameters requiring estimation. Finally, year were employed to estimate the dependency among the indices. As
the finalized predictive model and formula derivation of each index is shown in Formula (4.9), W x →2
xy represents the dependency weight of X
listed in Table 4. relative to Y when X is the control index; it also represents the effect of
The rates of data changes of the indices from 2009/Q2 to 2016/Q4 the rate of change in Y on the rate of change in X when X is the control
were calculated according to the time-series analysis of the dynamic index. Formula (4.10) indicates the effect of the rate of change in Y on
trends of each index at different time points. Thus, each pair of indices the rate of change in X relative to the index Z when X is the control
had the same calculation and comparison criteria in the subsequent index.
DNP analysis. The quarterly rate of change was calculated using the
Δx + Δy
following formula: W x2→ xy =
Δx (4.9)
T + T2 + T1
X(t ) = 3 where W x2→ xy represents the dependency weight of X relative to Y when
3 (4.6)
X is the control index; Δx represents the rate of change in X; Δy
where X(t ) represents the mean of the index X in quarter t; T1, T2 , and T3 represents the rate of change in Y.
respectively represent the first, second, and final months of a specified
Δx + Δy Δx + Δz
quarter. W2x → yz = /
Δx Δx (4.10)
ΔX(t ) = (Xt − Xt − 1)/ Xt − 1 (4.7)
where W2x → yz
represents the dependency weight of Y relative to Z when
where ΔX(t ) represents the rate of change in X in quarter t relative to the X is the control index; Δx , Δy , and Δz respectively represent the rates
preceding quarter; X(t ) represents the mean of X in quarter t; and Xt − 1 of changes in X, Y, and Z.
represents the mean of X in the preceding quarter. Table 5 lists the
quarterly rates of change in each index; the rates of change from 2009/ 4.5.1.3. Weights of the indices involving their internal dependency
Q2 to 2014/Q4 were converted from the data, whereas those from (wC ). Formula w1 × W2 = wc was employed to obtain the weights of
2015/Q1 to 2016/Q4 were projected using the prediction models. the indices with internal dependency (wc ). Finally, the extent of change
in the indices at different time points was used to calculate the changes
4.5. Verifying the DNPs in the weights of the indices over the 31 quarters (2009/Q2–2016/Q4).

There are two types of DNPs: data-oriented and expert-oriented. For 4.5.2. Verifying the expert-oriented DNP
the data-oriented DNP, the rate of change in each index was applied as To compare the effectiveness of data-oriented and expert-oriented
the basis for comparing the index pairs and calculating their weights; DNPs, a questionnaire survey was administered to experts. To maintain
for the expert-oriented DNP, the questionnaires were used to obtain consistency between the two phases of analysis of the questionnaire
scores from experts for each index. The results of the two types of DNP survey, the same interviewees were recruited for the DNP and FDM
were compared. questionnaire surveys. Because of the complexity in the questionnaires,
Regarding the dependency among the indices in the DNP, the FDM only one expert each from the industrial, governmental, and academic
questionnaire was employed to identify the interactive relationships sector was interviewed in the DNP survey. The expert-oriented DNP was
among the transportation sustainability indices with the closest fit divided into the following three steps:
(Fig. 4). The interactive relationships were then implemented in the
expert-oriented and data-oriented DNPs to obtain the weighting matrix 4.5.2.1. Weights of the indices without considering their internal
of the internal dependency among the indices (W2). dependency (w1). The experts were asked to use the revised version of

17
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

Table 5
Quarterly rates of change in each index (2009–2016).
Quarter Index A1 Index A2 Index B1 Index B2 Index C1 Index C2 Index C3

Land Consumption in Use of Public and Congestion Government Investment Accessibility of the Safety (Number of Health Impact
Transportation Non-motor Severity on Transportation Public Transportation Vehicle Accidents/ (Pollutant Standards
Vehicles 10,000 Vehicles) Index)

2009/Q2 −0.00014 0.016678 −0.00097 0.0001 0.002176 −0.35714 0.264544

2009/Q3 −0.00014 0.044736 0.001461 0.0001 0.132447 0.111111 −0.24504
2009/Q4 −0.00014 0.055548 0.007778 −0.00557 0.001655 0.00001 0.047339
2010/Q1 0.009523 −0.03808 0.005306 0.123318 −0.00209 0.4 0.828113
2010/Q2 0.009523 0.025127 −0.0024 0.001996 0.002708 −0.28571 −0.08525
2010/Q3 0.009523 0.021078 0.003367 0.000996 −0.00536 0.2 −0.1135
2010/Q4 0.009523 0.127379 0.0001 −0.00597 0.071742 0.0001 0.010333
2011/Q1 −0.00072 −0.02405 0.0001 −0.0991 0.037016 0.083333 0.30533
2011/Q2 −0.00072 0.00624 −0.00048 0.0001 0.005353 −0.30769 0.125244
2011/Q3 −0.00072 0.029716 0.001439 0.0001 −0.00358 0.0001 −0.09557
2011/Q4 −0.00072 0.037139 0.001437 0.062222 0.000767 0.222222 0.026097
2012/Q1 −0.00072 −0.00792 −0.00096 0.051255 0.062727 0.090909 0.617223
2012/Q2 −0.00072 −0.00575 −0.00335 0.00199 0.003125 −0.33333 0.137489
2012/Q3 −0.00072 0.021678 0.001441 0.001986 −0.00065 0.125 −0.12375
2012/Q4 −0.00072 0.075976 0.00001 0.004955 0.002135 0.444444 −0.17258
2013/Q1 −0.00022 −0.03882 0.0001 0.002959 −0.0019 −0.30769 0.90125
2014/Q1 −0.00022 0.010233 0.00048 0.0001 0.005882 0.111111 0.058119
2014/Q2 −0.00022 0.021588 −0.01103 0.000983 0.007302 0.3 −0.11379
2014/Q3 −0.00022 0.090505 −0.01794 0.003929 0.033937 −0.07692 0.141851
2014/Q4 0.000228 −0.02243 −0.01185 0.024462 0.015659 −0.25 0.450459
2015/Q1 0.000228 −0.00671 −0.00999 0.0001 0.004436 0.222222 0.015621
2015/Q2 0.000228 0.006151 −0.00706 0.0001 −0.00365 0.090909 −0.02642
2015/Q3 0.000228 0.040128 0.0001 0.030564 0.042099 0.916667 −0.03836
2015/Q4 0.001971 0.119708 −0.00203 0.0470 0.015122 −1.8029 0.006177
2016/Q1 0.002826 0.017684 −0.00088 −0.00146 0.009373 0.019165 0.076426
2016/Q2 0.002562 0.010641 0.0001 0.0001 0.011091 0.023055 −0.02353
2016/Q3 0.002353 0.001858 0.0001 0.0001 0.011833 0.015753 0.0026
2016/Q4 0.002188 0.017705 0.0001 0.0001 0.0121 0.009308 0.017395

Note: All the values from 2009/Q2 to 2014/Q4 were calculated from the actual data.

Fig. 4. Internal dependency among the indices.

18
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

the 1–9 scale created by Saaty (1990) to compare the seven indices in The weight of Index A1 remained relatively stable and was the fourth or
pairs, thereby obtaining the weightings of the indices without fifth highest for most of the study period. The weight of Index A2
considering their internal dependency (w1). Only those responses with gradually increased from third highest in 2009 to the highest in 2016.
consistency ratios lower than 0.1 in the matrix of a paired comparison The weight of Index B1 remained fifth or sixth highest overall. Index
were considered to satisfy the consistency test. B2, which maintained the highest weight in 2009–2013 and was third
highest in 2016, was a vital index for transportation sustainability.
4.5.2.2. Index internal dependency weighting matrix (W2 ). The experts Index C1 maintained the second highest weight throughout the study
assessed how each control index was affected by the other indices that period. The weight of Index C2 decreased to sixth before recovering to
exhibited internal dependency according to the extent of their fourth. Index C3 maintained the lowest weight throughout the study
influence. The following two conditions must be fulfilled: (a) the self- period with no significant change in its sequential order.
importance of each index was equal (importance = 1); (b) when a To further compare the results of the expert-oriented and data-or-
control index was present, the control index has higher importance than iented DNP, the data-oriented DNP calculation results from 2009 to
the other indices do (importance > 1). Although internal dependency 2016 were integrated into the 2009, 2013, and 2016 data through their
was observed among the indices, the self-importance of the control geometric means. Subsequently, the comparison results were organized
index was higher than those of the other indices. The level of internal into a weight sequence chart (Table 10.) and a dynamic index trend
dependency among the indices (W2 ) was calculated according to the comparison chart (Table 11).
FDM questionnaire survey results. Regarding the sequencing results shown in Table 10., Index B2 was
the most critical index in both the data-oriented and expert-oriented
4.5.2.3. Weights of the indices involving their internal dependency DNPs, and Index A2 also exhibited considerable importance. By con-
(wC ). The formula w1 × W2 = wc was used to calculate the weights of trast, Index C3 was regarded as the least critical index. The difference
the indices with internal dependency (wc ). Finally, the extent changes in between the two DNPs was the most significant for Index C1, the weight
the indices at different time points was used to calculate the changes in of which was fifth highest in the data-oriented DNP and second highest
the weights of the indices at three time points, namely in 2009, in 2013, in the expert-oriented DNP.
and in 2016. Significant differences were observed between the two DNPs in their
prediction of future trends in the indices. As revealed in Table 11, the
weight of Index A1 increased before decreasing in the expert-oriented
4.5.3. Comparing the empirical results of the two DNPs
DNP, but decreased slightly before increasing in the data-oriented DNP.
Table 6 and 7 and Fig. 5 illustrate the changes in the weights of the
The weight of Index A2 continually increased in the expert-oriented
transportation sustainability indices in the data-oriented DNP. Land
DNP, but decreased after peaking in the data-oriented DNP. The weight
consumption in transportation (Index A1) remained the lowest-
of Index B1 remained relatively steady in the expert-oriented DNP, but
weighted from 2009 to 2016, with no significant change in its trend.
it decreased substantially before increasing in the data-oriented DNP.
The weight of the use of public and non-motor vehicles (Index A2)
The weight of Index B2 continually decreased in the expert-oriented
gradually increased before progressively decreasing; the weight was the
DNP, but continually increased in the data-oriented DNP. The weight of
third to fifth highest. The weight of congestion severity (Index B1)
Index C2 increased slightly before decreasing and remaining relatively
gradually decreased before progressively increasing; this weight was
constant in the expert-oriented DNP, but decreased substantially before
the second to fifth highest. The weight of government investment on
increasing in the data-oriented DNP. The weight of Index C2 exhibited a
transportation (Index B2) gradually increased and remained the highest
profound decrease before increasing in the expert-oriented DNP, but it
from 2009 to 2016. The weight of government investment in trans-
increased significantly before decreasing in the data-oriented DNP. Fi-
portation (Index C1) gradually decreased before progressively in-
nally, the weight of Index C3, which exhibited the smallest difference
creasing; this weight was the third to fifth highest, exhibiting the si-
between the two DNPs, remained relatively constant in both the DNPs.
milar trend to Index B1. The weight of safety (Index C2) gradually
However, some slight differences could still be identified. The weight of
increased before progressively decreasing; this weight was the second
Index C3 gradually increased in the expert-oriented DNP, but slowly
to fifth highest for the majority of the period. The weight of health
decreased in the data-oriented DNP.
impact (Index C3) gradually decreased, remaining the second lowest
In summary, the expert-oriented and data-oriented DNPs were si-
throughout the study period. To further clarify the trends and compare
milar in parts of the weighting sequences and trend estimation, but
them to the results of the conventional DNP, the quarterly weight
significant differences were observed in most of the index analysis re-
changes were converted to annual weight changes, as shown in Table 8
sults. This was attributed to subjective opinions of the experts, who may
and Fig. 6.
have overlooked the complexity of the problems facing Taipei City. This
On the expert-oriented DNP, the questionnaire responses were used
might also be because the data selected in this study were not ideal for
to calculate the weight changes in the transportation sustainability in-
estimation. Consequently, some of the data could not thoroughly reflect
dices for Taipei City according to the responses of each expert. The
the conditions as predicted by the experts; some of the indices involve
geometric mean of the responses from the experts of the industrial,
numerous data and might be affected by factors that are difficult to
governmental, and academic domains was calculated for the equal
quantify. However, because the goal of this study was to assess the
considerations of the three domains, thereby revealing the dynamic
transportation sustainability indices and strategies for attaining high
weight changes and sequential order, as shown in Table 9 and Fig. 7

Table 6
Quarterly changes in the weights of the indices in the data-oriented DNP (2009Q2–2012Q4).
wc 2009 Q2 2009 Q3 2009 Q4 2010 Q1 2010 Q2 2010 Q3 2010 Q4 2011 Q1 2011 Q2 2011 Q3 2011 Q4 2012 Q1 2012 Q2 2012 Q3 2012 Q4

A1. 0.001 0.001 0.031 0.033 0.014 0.015 0.030 0.056 0.001 0.005 0.058 0.023 0.004 0.005 0.003
A2. 0.187 0.130 0.128 0.106 0.235 0.194 0.192 0.084 0.233 0.060 0.238 0.063 0.232 0.157 0.218
B1. 0.101 0.197 0.203 0.139 0.067 0.094 0.267 0.166 0.073 0.212 0.054 0.184 0.071 0.121 0.087
B2. 0.290 0.275 0.215 0.225 0.316 0.307 0.275 0.191 0.307 0.222 0.262 0.208 0.303 0.280 0.306
C1. 0.100 0.141 0.238 0.168 0.068 0.092 0.200 0.182 0.069 0.205 0.111 0.176 0.072 0.125 0.088
C2. 0.228 0.149 0.098 0.200 0.249 0.222 0.023 0.196 0.251 0.148 0.255 0.188 0.250 0.210 0.239
C3. 0.093 0.106 0.088 0.130 0.050 0.075 0.012 0.125 0.066 0.147 0.023 0.158 0.067 0.101 0.058

19
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

Table 7
Quarterly changes and sequential order of the weights of the indices in the data-oriented DNP (2013Q1–2016Q4).
wc 2013Q1 2013Q2 2013Q3 2013Q4 2014Q1 2014Q2 2014Q3 2014Q4 2015Q1 2015Q2 2015Q3 2015Q4 2016Q1 2016Q2 2016Q3 2016Q4

A1. 0.001 0.001 0.007 0.020 0.017 0.009 0.013 0.012 0.009 0.017 0.018 0.035 0.019 0.019 0.035 0.033
A2. 0.084 0.206 0.228 0.122 0.120 0.296 0.241 0.300 0.308 0.079 0.149 0.236 0.132 0.150 0.166 0.126
B1. 0.160 0.098 0.073 0.174 0.142 0.024 0.062 0.037 0.021 0.190 0.167 0.129 0.214 0.236 0.235 0.223
B2. 0.244 0.304 0.311 0.284 0.247 0.329 0.318 0.316 0.324 0.257 0.306 0.321 0.300 0.298 0.312 0.277
C1. 0.160 0.086 0.078 0.162 0.148 0.031 0.072 0.029 0.025 0.165 0.120 0.096 0.152 0.170 0.160 0.155
C2. 0.198 0.230 0.242 0.146 0.198 0.290 0.242 0.294 0.311 0.164 0.161 0.161 0.113 0.078 0.061 0.100
C3. 0.152 0.074 0.061 0.092 0.128 0.021 0.053 0.012 0.002 0.128 0.079 0.023 0.070 0.050 0.032 0.087

Fig. 5. Line chart showing the quarterly changes in the weights of the indices in Fig. 6. Line chart showing the annual changes in the weights of the indices in
the data-oriented DNP. the data-oriented DNP.

livability in Taipei City through the use of the data-oriented DNP, the Table 9
follow-up suggestions on sustainable transportation strategies were still Annual changes and sequential order of the index weights according to the
based on the data-oriented DNP analysis results. expert opinions.
wC 2009 Order 2013 Order 2016 Order
4.6. Sustainable transportation strategies to fulfill the livability of Taipei
City A1. Land Consumption in 0.1172 5 0.1600 4 0.1431 4
Transportation
A2. Use of Public and Non- 0.1427 3 0.1645 3 0.2002 1
The sustainable transportation strategies that fulfill the livability of
motor Vehicles
Taipei City were formulated according to the analysis results. The or- B1. Congestion Severity 0.1133 6 0.1279 5 0.1164 6
ientations of the sustainable transportation strategies in other cities B2. Government Investment 0.2723 1 0.2326 1 0.1696 3
were reviewed as references for revising the transportation strategies in in Transportation
Taipei. Subsequently, the dynamic changes in the index weights were C1. Accessibility of Public 0.1963 2 0.2158 2 0.1882 2
Transportation
analyzed to generalize the future developmental trend of the trans-
C2. Safety 0.1289 4 0.0729 6 0.1391 5
portation in Taipei City. Finally, discussions and suggestions were made C3. Health Impact 0.0290 7 0.0260 7 0.0431 7
on the 2020 White Paper on the Transportation Policies in Taipei.

4.6.1. Referencing urban transportation strategies involved six primary strategic goals: support economic and population
To properly and effectively develop urban transportation, city growth, strengthen the quality of daily living, improve safety and
governments have proposed transportation plans that are suitable for privacy, improve transportation accessibility, reduce the effect of
their respective cities, formulated the vision of those plans, and ex- transportation on climate change and improve transportation flex-
tended their durations to 2020, 2040, or even into the second half of the ibility, and support the 2012 London Olympics and Paralympics. In
century. The transportation dimension of the 2010 Greater London Plan addition, developmental strategies were established to support the six

Table 8
Annual changes and sequential order of the weights of the indices in the data-oriented DNP (2009–2016).
wc 2009 Order 2010 Order 2011 Order 2012 Order 2013 Order 2014 Order 2015 Order 2016 Order

A1 0.011 7 0.023 7 0.030 7 0.009 7 0.007 7 0.012 7 0.019 7 0.026 7

A2 0.148 5 0.182 2 0.153 3 0.167 3 0.160 3 0.239 3 0.193 3 0.143 4
B1 0.166 2 0.141 4 0.126 5 0.115 4 0.126 4 0.066 5 0.126 4 0.226 2
B2 0.260 1 0.280 1 0.245 1 0.274 1 0.286 1 0.302 1 0.302 1 0.296 1
C1 0.159 3 0.132 5 0.141 4 0.115 5 0.121 5 0.070 4 0.101 5 0.159 3
C2 0.158 4 0.173 3 0.212 2 0.222 2 0.204 2 0.255 2 0.199 2 0.088 5
C3 0.095 6 0.066 6 0.090 6 0.095 6 0.094 6 0.053 6 0.058 6 0.059 6

20
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

assessment indices.
In response to climate change, Santos and Ribeiro (2013) proposed
the following transportation strategies for Rio de Janeiro: (a) expand
the railway and subway network; (b) expand the bus rapid transit
system; (c) improve sustainable bio-fuel; (d) enforce bus plans; (e)
promote the use of electric vehicles; (f) implement light vehicle in-
spection and maintenance plans; and (g) formulate vegetable oil and
biodiesel reuse plans. Shiau (2012) formulated 15 sustainable trans-
portation strategies for Taipei City and determined their priority. In
particular, improving the accessibility of non-motorized transportation
modes, the use of biodiesel and bio-gasoline, demand responsive
transport services, and accessibility of transportation to elderly people
and people with disabilities are four of the five highest-priority strate-
gies.
In summary, most transportation strategies formulated worldwide
have involved the following aspects: (a) accessibility, promotion, and
construction of public transportation; (b) friendliness, safety, and
quality of daily living; (c) traffic management; (d) maintenance of re-
Fig. 7. Line chart showing the time-dependent weight changes in the indices in
the expert-oriented DNP.
lated assets; (e) protection of natural resources; (f) responses to climate
change; (g) improvement of system flexibility. Some strategies have
been associated with the disclosure of government information, land
goals. The Transportation 2040 Plan, established to improve land use use and urban plans, or master land use plans. Similarly, most of the
and transportation coordination in Puget Sound, involves the following contents of the 2020 White Paper on the Transportation Policies in Taipei
practical design principles: (a) promote mixed land use; (b) arrange focused on public transportation, friendliness, safety, quality of daily
residential areas to promote high-density residential growth; (c) con- living, and traffic management. However, few strategies on maintaining
nect neighborhoods with streets, alleys, and sidewalks; (d) integrate the assets, responding to climate change, disclosing government informa-
activity spaces of surrounding communities; (e) establish semipublic or tion, and linking transportation plans with urban development plans
public high-capacity transit stations near the centers of specified cities; have been proposed by the Taipei City Government. The disclosure of
(f) design plans for pedestrians and bicycles; (g) provide citizens with government information has been regarded only as a subtopic in smart
open spaces; (h) manage parking spaces; (i) improve the advantages of transportation. The association of transportation plans with urban
street parking; and (j) reduce the negative effects of parking. The planning is barely notable in strategies for constructing an environment
Southeast Michigan Council of Governments established their Principles that is comfortable, obstacle-free, and beneficial for developing public
Guiding Development of the 2040 Regional Transportation Plan, which are transportation. No consideration has been given to maintaining assets
described as follows: (a) emphasize construction on numerous regional and responding to climate change.
assets and provide strategic investments in local- and regional-scale
projects under finite financial resources; (b) disclose information per- 4.6.2. Future contexts and strategies for the transportation in Taipei City
taining to transportation to the public; (c) improve the comprehen- The goal of this study was to assess the transportation sustainability
siveness of transportation systems; (d) strengthen various modes of indices and strategies for Taipei City to achieve the goal of livability.
travel and the connection among local residents, communities, and Therefore, the follow-up suggestions of transportation planning were
global society; (e) enhance the flexibility and adaptability of transpor- based on the data-oriented DNP analysis results. According to the
tation in response to environmental dynamics; (f) provide various aforementioned index analysis results, the future conditions of trans-
groups with a cooperation framework; and (g) coordinate the plan with portation in Taipei City were analyzed, and suggestions were proposed
other projects such as macroeconomic development strategies, sus- on the 2020 White Paper on the Transportation Policies in Taipei. The
tainability frameworks, residential strategies, environmental plans, and future conditions regarding the transportation plans in Taipei were
green infrastructures. The goals of the 2040 Regional Transportation analyzed as follows:
Plan by the Houston–Galveston Area Council were aimed at improving
safety, reducing congestion, safeguarding ideal asset management and (a) No significant changes in the land consumption and use in trans-
operations, strengthening regional economic competitiveness, and portation were observed. This was ascribed to the status of Taipei
conserving and protecting natural and cultural resources. These goals City as a high-density development zone, with land use ap-
correspond to four strategies: improving system management and op- proaching saturation. Although the Taipei MRT network continues
erations, improving the states of ideal maintenance, expanding the to under development, most developments have occurred in the
multimodal network, and coordinating development. Finally, the ef- New Taipei City areas surrounding Taipei City. Nearly all of the
fectiveness of the strategies is examined through the use of relevant

Table 10
Weight sequence results of the data-oriented and expert-oriented DNPs.
wC 2009 2013 2016

Data Conventional Data Conventional Data Conventional

A1. Land Consumption in Transportation 7 5 7 4 7 4

A2. Use of Public and Non-motor Vehicles 3 3 3 3 3 1
B1. Congestion Severity 4 6 4 5 2 6
B2. Government Investment in Transportation 1 1 1 1 1 3
C1. Accessibility of Public Transportation 5 2 5 2 5 2
C2. Safety 2 4 2 6 4 5
C3. Health Impact 6 7 6 7 6 7

21
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

Table 11
Weighting sequences in the data-oriented and expert-oriented DNPs.

land in Taipei City has been used in planning and construction; throughout the specified years; not only has its weight changed
undeveloped zones have not change substantially in their number considerably, but it has had also profound effected on the other
because of problems in acquiring reserved land. Therefore, little index weights.
concern is required for Index A1 (land consumption in transporta- (d) The weight of Index C2 (safety) has been one of the highest in the
tion) in current transportation plans for Taipei City. overall weighting sequence. Although its weight has dropped more
(b) The weight of Index A2 (use of public and non-motor vehicles) recently, it has remained highly crucial for transportation planning
peaked in 2013. In 2009–2013, numerous MRT lines were opened, in Taipei City. This reveals that the urban transportation safety
and the construction of public bicycle stations was expanded, ex- pertaining to roads, sidewalks, and bicycle lanes must continue to
ceeding 100 in 2013. Hence, the rate of the use of public and non- receive consideration in future transportation strategies for Taipei
motor vehicles peaked in that year. Although the weight of A2 City. Index C3 (health impact) has remained among the indices with
decreased slightly in 2016, this index remains among the most the lowest weights. Because the applied data corresponding to the
prioritized indices. Therefore, Index A2 should receive considerable index was the air pollution index, the air quality in Taipei City was
attention in transportation planning. Substantially associated to considered stable and acceptable. Therefore, improving this index
Index A2 is Index C1 (accessibility of the public transportation). would not substantially affect the overall quality of transportation
Therefore, Index C1 should also be highly prioritized in transpor- in Taipei City. This index does not require attention in the short
tation planning. According to the status of Index A1, the construc- term.
tion of additional MRT stations may be limited by land conditions.
Therefore, decision makers should increase the use of public and
non-motor vehicles through other dimensions and improve their 4.6.3. Suggestions for the 2020 white paper on the Transportation Policies
accessibility without increasing land use. in Taipei
(c) The weight of Index B1 (congestion severity) continually increased The general orientation of the transportation strategies for Taipei
over time. This is attributable to the increasing prevalence of public City was clarified through the situational analysis based on the index
and non-motor vehicles. Index B2 (government investment in weighting results. The analysis results correspond to the suggestions for
transportation) was consistently the highest-weighted index the urban transportation planning in Taipei City in the 2020 White
Paper on the Transportation Policies in Taipei.

22
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

Fig. 8. Development topics and execution strategies in the 2020 White Paper on Transportation Policies in Taipei. Source: Department of Transportation, Taipei City
Government.

This white paper lists sustainable transportation, affinity and hu- (a) Establish a comfortable and obstacle-free transportation environ-
manism, and livability and safety as the visions of development. ment
Corresponding to the visions are four specific development goals: in- (b) Increase the usage rate of public transportation
crease the usage rate of environmentally friendly vehicles to 70%, fully (c) Examine transportation services and fee subsidy mechanisms
implement low-pollution public transportation vehicles in the city, (d) Create a safe and accessible transportation system
strengthen obstacle-free environments, and reduce the rate of traffic (e) Promote smart transportation systems
accidents by 3% annually. Numerous execution strategies (Fig. 8) were (f) Create an environment suitable for developing the public trans-
formulated on the basis of the following dimensions: portation network

23
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

(g) Conserve energy and reduce carbon emissions (i.e., fossil fuel usage (a) Aggregating the transportation sustainability indices: Previous
and motor vehicle emissions) studies on livable cities and sustainable transportation (Section 2)
(h) Construct an environmentally friendly transportation environment were generalized to derive 18 indices along three dimensions (en-
vironmental, economic, and social).
The four development goals of the white paper correspond to the (b) Selecting the indices applicable for attaining the livability of
results of the analysis of the index weights. Index A2 (use of public and Taipei City: From the preliminary list of transportation sustain-
non-motor vehicles) accords with the goal of introducing low-pollution ability indices that satisfy the principles of a livable city, the first-
public transportation vehicles, and Index C2 (Safety) accords with the phase FDM expert questionnaire survey was performed. Six experts
goal of reducing the rate of traffic accidents. In the index weight dy- with industrial, governmental, and academic backgrounds in
namic analysis, both of these indices were among those with the highest transportation were invited to evaluate the indices. The threshold
priority. Therefore, the development goals of introducing low-pollution value for FDM selection was 7.30; seven critical indices were above
public transportation vehicles and reducing the rate of traffic accidents this threshold.
should be prioritized over increasing the usage rate of environmentally (c) Data-oriented DNP results: The rates of changes in the indices
friendly vehicles and improving obstacle-free environments. were implemented in the DNP to obtain the quarterly weight
On the overall transportation development in the city, increasing changes in the indices in 2009–2016. The weights of Index A1 were
the usage rate of public transportation, examining the transportation 0.1172, 0.1600, and 0.1431 in 2009, 2013, and 2016, respectively.
services and fee subsidy mechanisms, creating a safe and easy-to-un- The weight non-significantly increased before decreasing over the
derstand transportation system, conserving energy and reducing carbon years. Therefore, the function of the index is d1 t 2 + d2 t + d3 (d1 is
emissions, creating an environment suitable for developing public negative). The index had the lowest weight of the seven selected
transportation, and constructing an environmentally friendly trans- indices. The corresponding weights for Index A2 were 0.1427,
portation environment are consistent with the assessment dimensions 0.1645, and 0.2002. This index was initially among the top three
generalized in this study and should be prioritized as strategic dimen- indices but became the top-priority index in 2016, increasing in a
sions for developing sustainable transportation in Taipei City. linear manner. Therefore, the function of the index is a1 t + a2 (a1 is
Regarding establishing comfortable and obstacle-free transportation positive). The corresponding weights of Index B1 were 0.1133,
environments and promoting smart transportation systems, two related 0.1279, and 0.1164. No significant changes were observed in the
indices were disregarded in the FDM phase of the index assessment, weight of the index or its priority. Therefore, the constant a, which
because they are the lowest-priority dimensions for the sustainable did not change over time, was used to interpret the index. The re-
transportation development. For easy distinction, Fig. 8 displays the spective weights of Index B1 were 0.2723, 0.2326, and 0.1696.
strategic dimensions and strategies consistent with the weight analysis Because the weight of this index decreased steadily and linearly
results colored in yellow and orange (the other dimensions and stra- over time, the function of the index is a1 t + a2 (a1 is negative). The
tegies are temporarily disregarded). overall priority of this index was the third-highest in 2016. The
On the correspondence between the indices and strategies (Fig. 9), corresponding weights of Index C1 were 0.1963, 0.2158, and
managing parking spaces, establishing a fee-payment mechanism for 0.1882. No significant changes were observed in the weight of the
vehicles entering specific regions, and externalizing parking grids on index or its priority. Therefore, the constant a, which did not
the sides of the main roads correspond to Index B1 (congestion se- change over time, was used to interpret the index. The respective
verity). Increasing the number of bicycle lanes and parking facilities in weights of Index C2 were 0.1289, 0.0729, and 0.1391. This index
the city, promoting the density of public transportation stations, and was one of three highest-priority indices over the study period.
promoting public-transportation-oriented development in newly de- Because the weight of this index initially decreased and then in-
veloped regions correspond to the highly weighted Index C1 (accessi- creased over time, the function of this index is d1 t 2 + d2 t + d3 (d1 is
bility of public transportation) and must be prioritized in the devel- positive). The corresponding weights of Index C3 were 0.0290,
opment strategies. Strengthening the demand for response services, 0.0260, and 0.0431. No significant changes were identified in the
rationalizing public transportation fees and sustainably managing weight of the index and its priority. Therefore, the constant a,
public transportation, and providing comfortable pedestrian spaces which did not change over time, was used to interpret the index.
correspond to the highly weighted Index A2 (use of public and non- (d) Comparison of the data- and expert-oriented DNPs: The se-
motor vehicles), and promoting road traffic safety corresponds to the quencing results of the two DNPs show that Index B2 was the most
similarly weighted Index C2 (safety). These strategies must be prior- critical index in both DNPs, Index A2 also exhibited considerable
itized. Integrating regional transportation networks, reallocating road importance, and Index C3 was the least critical. The most notable
spaces, and promoting low-pollution public vehicles correspond to the difference between the two DNPs was in Index C1, which was the
low-weight Indices A1 (land consumption in transportation) and C3 fifth highest-weighted in the data-oriented DNP but second in the
(health impact) and are thus low-priority strategies. Popularizing low- expert-oriented DNP. Significant differences between the two DNPs
floor buses, humanizing transportation facilities, expanding value- were also noted regarding their prediction of the future trends of
added information applications, integrating and managing traffic in- the indices. As shown in Table 11, the weight of Index A1 increased
formation, and providing convenient public information services do not before decreasing in the expert-oriented DNP, but it decreased
correspond to any of the indices examined in this study. Therefore, slightly before increasing in the data-oriented DNP. In the expert-
these strategies are lowest-priority in promoting sustainable transpor- oriented DNP, the weight of Index A2 continually increased, but it
tation in Taipei City. decreased after increasing in the data-oriented DNP. The weight of
Index B1 remained relatively constant in the expert-oriented DNP,
but it decreased substantially before increasing in the data-oriented
5. Policy implication and conclusion DNP. In the expert-oriented DNP, the weight of Index B2 con-
tinually decreased, but the opposite trend was observed in the data-
This section summarizes the analysis results, namely the pre- oriented DNP. The weight of Index C2 slightly increased before
liminary generalization and selection of the indices, results of the data- decreasing or remaining relatively constant in the expert-oriented
oriented DNP, and comparison between the data- and expert-oriented DNP, but a substantial decrease was observed before an increase in
DNPs. In addition, from the data-oriented DNP results, suggestions for the data-oriented DNP. The weight of Index C2 profoundly de-
the Taipei City Government regarding the white paper on transporta- creased before increasing in the expert-oriented DNP, yet the op-
tion policies in Taipei City are offered. posite was observed in the data-oriented DNP. Finally, the weight of

24
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

Fig. 9. Priority sequence of the execution strategies in the 2020 White Paper on Transportation Policies in Taipei.

Index C3, which exhibited the smallest difference between the two introducing low-pollution public transportation and reducing the
DNPs, remained constant throughout the study period in both the rate of traffic accidents should be regarded as the most critical
DNPs. However, some slight differences were identified. The weight goals.Regarding the execution of the strategies, the management of
of Index C3 gradually increased in the expert-oriented DNP, but the parking spaces, establishment of a fee-payment mechanism for the
opposite was observed in the data-oriented DNP. vehicles entering specific regions, and externalization of parking
(e) Suggestions for the white paper on the transportation policies grids to the sides of the main roads correspond to Index B1 (con-
in Taipei City according to the dynamic trends of the index gestion severity). Increasing the number of bicycle lanes and
weights: Regarding the four development goals listed in the 2020 parking facilities in the city, promoting the density of public
White Paper on the Transportation Policies in Taipei, the dynamic transportation stations, and focusing on the development of public
analysis of the index weights shows that Indices A2 and C2 are transportation systems in newly developed regions correspond to
among the highest-priority indices. Therefore, the two goals of the highly weighted Index C1 (accessibility of public

25
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

transportation) and must be prioritized as development strategies. Acknowledgements

Strengthening the demand for response services, rationalizing
public transportation fees, sustainably managing public transpor- The authors would like to thank anonymous referees and the Editor
tation, and providing comfortable pedestrian spaces correspond to of the journal for constructive comments and suggestions. The authors
the highly weighted Index A2 (use of public and non-motor also wish to acknowledge Wei-En Hsu for his participation in data
Vehicles), and promoting road traffic safety corresponds to Index collection and analysis activities. Finally, the authors would like to
C2 (safety), which was similarly weighted. These strategies must be express their appreciations to Ministry of Science and Technology
prioritized. Integrating regional transportation networks, re- (MOST) of Taiwan for the support of the projects 104-2410-H-305-077-
allocating road spaces, and promoting low-pollution public vehicles MY3, 106-2410-H-305-057-MY3 and 107-NTPU_A-H&E-143-001. The
correspond to the low-weight Indices A1 (land consumption in views expressed are those of the authors and do not represent the of-
transportation) and C3 (health impact), both of which are low- ficial policy or positions of the MOST.
priority strategies. Popularizing low-floor buses, humanizing
transportation facilities, expanding value-added information ap- References
plications, integrating and managing traffic information, and pro-
viding convenient public information services did not correspond to Ahmad, N., & Mehmood, R. (2016). Enterprise systems and performance of future city
any of the indices investigated in this study. Therefore, these stra- logistics. Production Planning & Control, 27(6), 500–513.
Alam, F., Mehmood, R., Katib, I., Albogami, N. N., & Albeshri, A. (2017). Data fusion and
tegies should be considered the lowest-priority strategies in pro- IoT for smart ubiquitous environments: A survey. IEEE Access, 5, 9533–9554.
moting the sustainable transportation in Taipei City. The 2020 Alazawi, Z., Alani, O., Abdljabar, M. B., Altowaijri, S., & Mehmood, R. (2014). A smart
White Paper on the Transportation Policies in Taipei focused on public disaster management system for future cities. Paper presented at the Proceedings of the
2014 ACM international workshop on Wireless and mobile technologies for smart cities,
transportation, friendliness, safety, quality of daily living, and Philadelphia, Pennsylvania, USA.
traffic management, similar to urban development strategies in Arfat, Y., Aqib, M., Mehmood, R., Albeshri, A., Katib, I., Albogami, N., et al. (2017).
most cities. However, the Taipei City Government has proposed few Enabling smarter societies through mobile big data fogs and clouds. Procedia
Computer Science, 109, 1128–1133.
strategies on maintaining assets, responding to climate change, Awasthi, A., & Chauhan, S. S. (2011). Using AHP and Dempster–Shafer theory for eval-
disclosing government information, and associating transportation uating sustainable transport solutions. Environmental Modelling & Software, 26(6),
planning with urban development planning. The disclosure of 787–796.
Bagchi, M., & White, P. R. (2005). The potential of public transport smart card data.
government information has been regarded as only a subtopic in the
Transport Policy, 12(5), 464–474.
development of smart transportation systems. The association of Batty, M. (2008). The dilemmas of physical planning. Environment and Planning B:
transportation planning with urban planning is barely notable in Planning and Design, 35(5), 760–761.
strategies aimed at constructing an environment that is comfor- Batty, M. (2012). Building a science of cities. Cities, 29, S9–S16.
Batty, M. (2013). Big data, smart cities and city planning. Dialogues in Human Geography,
table, obstacle-free, and beneficial for developing public transpor- 3(3), 274–279.
tation. No consideration has been attached to maintaining assets Batty, M. (2017). Cities in disequilibrium. In J. Johnson, A. Nowak, P. Ormerod, B.
and responding to climate change. Rosewell, & Y.-C. Zhang (Eds.). Non-equilibrium social science and policy: Introduction
and essays on new and changing paradigms in socio-economic thinking (pp. 81–96).
Cham: Springer International Publishing.
Here, we must make a brief declaration that this study does not Batty, M., & Marshall, S. (2017). Thinking organic, acting civic: The paradox of planning
involve any emerging technology or new hardware architecture (Alam, for Cities in Evolution. Landscape and Urban Planning, 166, 4–14.
Bettencourt, L. M., Lobo, J., Helbing, D., Kühnert, C., & West, G. B. (2007). Growth,
Mehmood, Katib, Albogami, & Albeshri, 2017; Arfat et al., 2017; innovation, scaling, and the pace of life in cities. Proceedings of the National Academy
Rashid, Royston, Gary, Patrick, & Mukesh, 2017), nor does it involve of Sciences, 104(17), 7301–7306.
any personal information from emerging technology users. The focus of Bettencourt, L., & West, G. (2010). A unified theory of urban living. [Comment]. Nature,
467, 912.
this study is to return to the essence of urban development with dy- Bibri, S. E., & Krogstie, J. (2017). The core enabling technologies of big data analytics and
namic perspective that toward realistic urban situations. And, for context-aware computing for smart sustainable cities: A review and synthesis.
achieving ultimate goal of urban sustainability and urban livability [journal article]. Journal of Big Data, 4(1), 38.
Black, W. R. (1996). Sustainable transportation: A US perspective. Journal of Transport
(Yigitcanlar and Teriman, 2015), we expect to formulate a more
Geography, 4(3), 151–159.
transparent and realistic urban development strategies that incorporate Boyd, d., & Crawford, K. (2012). Critical questions for big data. Information,
only few concise tools and materials, including Fuzzy Delphi method, Communication & Society, 15(5), 662–679.
extendible open data (i.e., big data), auto-regressive integrated moving Chan, Shin-Liang, Wey, Wann-Ming*, & Chang, Pin-Huai (2014). Establishing Disaster
Resilience Indicators for Tan-sui River Basin in Taiwan. Social Indicators Research,
average model and Analytic Network Process. Unlike the previous ANP 115(1), 387–418.
or AHP studies that focus on converting qualitative information into Chang, Y.-H., Wey, W.-M., & Tseng, H.-Y. (2009). Using ANP priorities with goal pro-
quantitative parameters (Wey & Chiu, 2013; Chan, Wey & Chang, gramming for revitalization strategies in historic transport: A case study of the
alishan forest railway. Expert Systems with Applications, 36(4), 8682–8690.
2014), in this study, we firstly attempt to integrate the dynamic changes Deakin, M., Curwell, S., & Lombardi, P. (2002). Sustainable urban development: The
of extendible open data (big data) and ANP mechanisms to address framework and directory of assessment methods. Journal of Environmental Assessment
quantitative issues directly. In other words, our main innovation not Policy and Management, 04(02), 171–197.
Evans, P. (Ed.). (2002). Livable cities? Struggles for livelihood and sustainability. Berkeley:
only reveals the dynamic weight changes in data-oriented ANP, but also University of California Press.
expresses the dynamic concept of indices or alternatives that change Fidler, D., Olson, R., & Bezold, C. (2011). Evaluating a long-term livable communities
over time. Surely, the emerging technologies, industry-driven me- strategy in the U.S. Futures, 43(7), 690–696.
Floridi, L. (2012). Big data and their epistemological challenge. [journal article].
chanisms, intelligent systems or any ICT-based hardware (e.g., auto-
Philosophy & Technology, 25(4), 435–437.
nomic systems, self-driving vehicles, cloud computing, mobile big data Gal-Tzur, A., Grant-Muller, S. M., Kuflik, T., Minkov, E., Nocera, S., & Shoor, I. (2014).
fogs, internet of things, and distributed systems) that were not covered The potential of social media in delivering transport policy goals. Transport Policy, 32,
115–123.
in this study still worth to be considered in further to promote the three
Gans, H. J. (2006). Jane Jacobs: Toward an understanding of “Death and life of great
priority development goals (i.e., smart growth; sustainable growth, american cities”. City & Community, 5, 213–215.
inclusive growth) that set by European Union (Schlingensiepen et al., Håkonsson, T., & Carroll, T. (2016). Is there a dark side of Big Data – point, counterpoint.
2016). Finally, we reiterate that the policy implications of this study [journal article]. Journal of Organ Dysfunction, 5(1), 5.
Janssen, M., & Kuk, G. (2016). The challenges and limits of big data algorithms in
may offer a solution for current needs related to urban planning in the technocratic governance. Government Information Quarterly, 33(3), 371–377.
whole world, especially in developing cities, and may serve as a more Lai, S.-k., & Huang, J.-y. (2017). Theoretical foundation of a decision network for urban
transparent decision-making or selection basis for developing sustain- development. [journal article]. Frontiers of Information Technology & Electronic
Engineering, 18(8), 1033–1039.
able and livable urban life in near future. Laurence, P. L. (2006). The Death and life of urban design: Jane Jacobs, the rockefeller

26
W.-M. Wey, J.-Y. Huang Habitat International 82 (2018) 9–27

foundation and the new research in urbanism, 1955–1965. Journal of Urban Design, Soto, V., & Frias-Martinez, E. (2011). Robust land use characterization of urban land-
11(2), 145–172. scapes using cell phone data. The first workshop on pervasive Urban Applications (pp. 1–
Litman, T., & Burwell, D. (2006). Issues in sustainable transportation. International Journal 8). .
of Global Environmental Issues, 6(4), 331–347. Storey, V. C., & Song, I.-Y. (2017). Big data technologies and Management: What con-
Manovich, L. (2011). Trending: The promises and the challenges of big social data. In M. ceptual modeling can do. Data & Knowledge Engineering, 108, 50–67.
K. Gold (Ed.). Debates in the digital humanities (pp. 460–475). Minneapolis: The Suma, S., Mehmood, R., Albugami, N., Katib, I., & Albeshri, A. (2017). Enabling next
University of Minnesota Press. generation logistics and planning for smarter societies. Procedia Computer Science,
Mehmood, R., & Graham, G. (2015). Big data logistics: A health-care transport capacity 109, 1122–1127.
sharing model. Procedia Computer Science, 64, 1107–1114. Taki, H. M., Maatouk, M. M. H., Qurnfulah, E. M., & Aljoufie, M. O. (2017). Planning TOD
Naim, A., & Rashid, M. (2015). Enterprise systems: Are we ready for future sustainable with land use and transport integration: A review. Journal of Geoscience, Engineering,
cities. Supply Chain Management: International Journal, 20(3), 264–283. Environment, and Technology, 2(1), 84–94.
Newman, P., Beatley, T., & Boyer, H. (2009). Resilient cities: Responding to peak oil and Tsai, C.-W., Lai, C.-F., Chao, H.-C., & Vasilakos, A. V. (2015). Big data analytics: A survey.
climate change. Washington: Island Press. [journal article]. Journal of Big Data, 2(1), 21.
Newman, P., & Kenworthy, J. (2015). The end of automobile dependence: How cities are West, G. B. (2010). Integrated sustainability and the underlying threat of urbanization. In
moving beyond car-based planning. Washington, DC: Island Press. H. J. Schellnhuber, M. Molina, N. Stern, V. Huber, & S. Kadner (Eds.). Global sus-
Pelletier, M.-P., Trépanier, M., & Morency, C. (2011). Smart card data use in public tainability: A nobel cause (pp. 9–18). Cambridge: Cambridge University Press.
transit: A literature review. Transportation Research Part C: Emerging Technologies, West, G. (2017). Scale: The universal laws of growth, innovation, sustainability, and the pace
19(4), 557–568. of life in organisms, cities, economies, and companies. New York: Penguin Press.
Rashid, M., Royston, M., Gary, G., Patrick, H., & Mukesh, K. (2017). Exploring the in- Wey, W.-M. (2015). Smart growth and transit-oriented development planning in site se-
fluence of big data on city transport operations: A markovian approach. International lection for a new metro transit station in Taipei, Taiwan. Habitat International, 47,
Journal of Operations & Production Management, 37(1), 75–104. 158–168.
Reisi, M., Aye, L., Rajabifard, A., & Ngo, T. (2014). Transport sustainability index: Wey, W.-M., & Chiu, Y.-H. (2013). Assessing the walkability of pedestrian environment
Melbourne case study. Ecological Indicators, 43, 288–296. under the transit-oriented development. Habitat International, 38, 106–118.
Saaty, T. L. (1990). The analytic hierarchy process. Pittsburgh: RWS. Wey, W.-M., & Wu, K.-Y. (2007). Using ANP priorities with goal programming in resource
Saaty, T. L. (2007). Time dependent decision-making; dynamic priorities in the AHP/ allocation in transportation. Mathematical and Computer Modelling, 46(7), 985–1000.
ANP: Generalizing from points to functions and from real to complex variables. Wey, W.-M., Zhang, H., & Chang, Y.-J. (2016). Alternative transit-oriented development
Mathematical and Computer Modelling, 46(7–8), 860–891. evaluation in sustainable built environment planning. Habitat International, 55,
Santos, A. S., & Ribeiro, S. K. (2013). The use of sustainability indicators in urban pas- 109–123.
senger transport during the decision-making process: The case of Rio de Janeiro, Xing, Y., Horner, R. M. W., El-Haram, M. A., & Bebbington, J. (2009). A framework model
Brazil. Current Opinion in Environmental Sustainability, 5(2), 251–260. for assessing sustainability impacts of urban development. Accounting Forum, 33(3),
Schlingensiepen, J., Mehmood, R., & Nemtanu, F. C. (2015). Framework for an autonomic 209–224.
transport system in. Smart Cities, 15(5), 50. Yang, Y.-N. (2010). Financial Econometric with gretl free software. Taipei: COMPASS (in
Schlingensiepen, J., Nemtanu, F., Mehmood, R., & McCluskey, L. (2016). Autonomic Chinese).
transport management systems—enabler for smart cities, personalized medicine, Yigitcanlar, T., & Teriman, S. (2015). Rethinking sustainable urban development:
participation and industry grid/industry 4.0. In A. Sładkowski, & W. Pamuła (Eds.). Towards an integrated planning and development process. International Journal of
Intelligent transportation systems – problems and perspectives (pp. 3–35). Cham: Springer Environmental Science and Technology, 12(1), 341–352.
International Publishing. Zhan, D., Kwan, M.-P., Zhang, W., Fan, J., Yu, J., & Dang, Y. (2018). Assessment and
Shepherd, A., & Ortolano, L. (1996). Strategic environmental assessment for sustainable determinants of satisfaction with urban livability in China. Cities, 79, 92–101.
urban development. Environmental Impact Assessment Review, 16(4), 321–335. Zheng, Y., Liu, Y., Yuan, J., & Xie, X. (2011). Urban computing with taxicabs. Paper
Shiau, T.-A. (2012). Evaluating sustainable transport strategies with incomplete in- presented at the Proceedings of the 13th international conference on Ubiquitous computing,
formation for Taipei City. Transportation Research Part D: Transport and Environment, Beijing, China.
17(6), 427–432.

27