lessons in urban monitoring taken from sustainable and livable cities to better address the smart...

Technological Forecasting & Social Change xxx (2014) xxx–xxx

TFS-17940; No of Pages 12

Contents lists available at ScienceDirect

Technological Forecasting & Social Change

Lessons in urban monitoring taken from sustainable and livable cities tobetter address the Smart Cities initiative

Maria-Lluïsa Marsal-Llacuna a,⁎, Joan Colomer-Llinàs b,1, Joaquim Meléndez-Frigola b,2

a Department of Architecture and Urban Planning, Campus Montilivi, Politécnic Building III, Room 121, 17003 Girona, Spainb Institute of Informatics and Applications, Department of Electrical Engineering, Electronics and Automatics, Campus Montilivi, Politécnic Building IV,17003 Girona, Spain

a r t i c l e i n f o

⁎ Corresponding author. Tel.: +34 972 41 84 45.E-mail addresses: [email protected] (M.-L. Ma

[email protected] (J. Colomer-Llinàs), joaquim.m(J. Meléndez-Frigola).

1 Tel.: +34 972 41 8756.2 Tel.: +34 972 41 8883.

0040-1625/$ – see front matter © 2014 Elsevier Inc. Ahttp://dx.doi.org/10.1016/j.techfore.2014.01.012

Please cite this article as: M.-L. Marsal-Llabetter address the Smart Cities initiative, Te

a b s t r a c t

Article history:Received 1 October 2012Received in revised form 6 January 2014Accepted 21 January 2014Available online xxxx

In this paper we put forward two ideas for monitoring the Smart Cities initiative in a betterway.In developing the first idea, we study past and on-going initiatives in the field of sustainablecities and livable cities and their respective monitoring indicators to demonstrate that not onlyis a set of indicators needed for efficient monitoring, but also a final synthetic or aggregativeindex to visualize the initiative's achievements. Specifically, we propose the construction ofsynthetic indices using principal component analysis (PCA). The second idea attempts toanticipate the changes needed, especially with regard to data collection, to be introduced incurrent monitoring practices to assess a city's “smartness” accurately. We propose the use ofreal-time data instead of historical statistics as the basic information with which to construct aset of indicators to explain the initiative. A final index summarizing Smart Cities' real-time setof indicators is suggested in the conclusion.

© 2014 Elsevier Inc. All rights reserved.

Keywords:Urban monitoringUrban indicatorsUrban indicesSustainable citiesLivable citiesSmart Cities

1. Introduction: on the need for a synthetic index tovisualize the monitoring of urban strategies

It can be said that urban monitoring started with the EarthSummit of 1992, a pioneer event in promoting the role of citieson the road to sustainability. Out of the summit came theso-called Aalborg Charter, which began the assessment of cities'sustainability. By 1995, more than 1200 cities all over the worldhad ratified the charter. The main agreement reached by thesignatorieswas todrawup their ownLocal Agenda21, consistingof a set of indicators to monitor sustainability. From 1995 andover a period of approximately ten years a number of cities didthis. Since 2006, however, sustainability monitoring has de-creased considerably. In our opinion, the lack of a synthetic index

rsal-Llacuna),[email protected]

ll rights reserved.

cuna, et al., Lessons in uchnol. Forecast. Soc. Cha

summarizing the overall set of sustainability indicators is thereason behind LA 21's failure, a claim which is analyzed in thefollowing section.

Concurrentwith the start of the Local Agenda crisis, in 2005, anew set of global indicators appeared. Quality of life indicatorswere proposed as a way of assessing a city's livability. Thepromoter of this new urban monitoring was a private corpora-tion, Mercer, a human resources and related financial servicesconsultancy, with its headquarters in New York City. Mercer isthe world's largest human resources consultancy, operating inmore than 40 countries. Prior to launching its livabilitymonitoring indicators and index, the so called Mercer's Qualityof Living Reports, the firm was well-known for its ranking ofcities in terms of cost of living, and its list of the world's mostexpensive cities for expatriate employees. Mercer's quality of lifereports are published on a yearly basis for 221 major cities allover the world. All its surveys are updated yearly. BesidesMercer's there is another measure of livability acceptedworld-wide: The Economist Intelligence Unit's quality-of-lifeindex. Indicators and the final indexwere first calculated in 2005and included data from 111 countries. Unlike Mercer's, The

rban monitoring taken from sustainable and livable cities tonge (2014), http://dx.doi.org/10.1016/j.techfore.2014.01.012

http://dx.doi.org/10.1016/j.techfore.2014.01.012

mailto:[email protected]




http://www.sciencedirect.com/science/journal/00401625


2 M.-L. Marsal-Llacuna et al. / Technological Forecasting & Social Change xxx (2014) xxx–xxx

Economist index ranks both countries and cities, making use ofMercer's data for cities. The Intelligence Unit is an independentcompany within The Economist Group, a world leader in thefield of financial information. The unit provides tailored advisory,management and analysis services to companies. It is knownfor its e-readiness index (concerning the effective use of ICTtechnologies to boost countries' economies and welfare provi-sion) and democracy index (ameasure of the level of democracyin both UN and non-UN countries).

Although both Mercer's and The Economist's sets of indica-tors and final indices have a shorter history, they look morepromising than local agendas. Both summarize a combination ofsubjective life-satisfaction and objective quality of life indicatorsin an aggregated index which makes their well-known rankingsof countries and cities possible. The need of synthetic indices isnot only for comparison matters out of a certain ranking. In ouropinion, in today's era of information excess it is crucial tosynthesize the large amounts of available information in smallbut representative visualizations, and both Mercer and TheEconomist Intelligence Unit seem aware of that when drawingup their indices. A proof that quality of life indices are successfulis that middle-sized cities not included in these reports producetheir ownquality of life indicators and indices, and the number ofthem doing so is growing every year. This is exactly the oppositeof what has occurred with local agendas that assess a city'ssustainability. We take a deeper look at the success story ofquality of life indices in the second section of this paper.

In recent times the Smart Cities initiative has expanded allover the world. It was around 2009 when the concept beganto be globally understood as the target for any city to achieve,no matter what its size. The initiative developed out of theprevious experiences ofmeasuring environmentally friendly andlivable cities, embracing the concepts of sustainability andquality of life but with the important and significant addition oftechnological and informational components. Reliable indicatorsdo not yet exist to measure how “intelligent” cities are, andneither does a summarizing index, but interest in the initiative isgrowing and it will not be long before the worlds of academia,business and government start to take notice. Besides theproposal we make about the use of real-time data to develop aSmart Cities set of indicators – learning from the previous ex-periences of sustainable cities and livable cities – we proposethe elaboration of a synthetic indicator to summarize a city's“intelligence” or “smartness”. In the third section of this paperwe explain the different sources of real-time data available fordeveloping the latest generation of indicators for monitoring acity's smartness. The conclusions recap on the proposal for a finalsynthetic index of the smart city set of indicators, to make itpossible to easily visualize a city's steps towards “smartness”.

2. Analyzing past and on-going initiatives

2.1. Learning from the Sustainable City and the problems withthe LA 21 set of indicators

Local Agendas 21 were conceived of as a sustainabilityagenda based in the principles of urban ecology [1]. From theorigins of the sustainability movement, sustainable develop-ment was projected as an activity best generated andmost appropriate at a local scale. This localization of thephenomenon was central to the design of the agendas [2]. Their

Please cite this article as: M.-L. Marsal-Llacuna, et al., Lessons in ubetter address the Smart Cities initiative, Technol. Forecast. Soc. Cha

local character resulted in their application not only to big andmedium sized cities but also to small towns in non-developedcountries [3]. LA 21s have a proven capacity to enhance sus-tainability and the quality of the urban environment for thebenefit of citizens, but the road to sustainability through LAs hasonly been in existence for 10 years. In our opinion, the lack of asummarizing indicator giving information about a city's overallsustainability level, thereby allowing comparisons and rankingsbetween cities, has been their big mistake.

Since the beginning, LA 21 contained a standard set ofindicators with few variations when it was applied locally.We exemplify the LA 21 set of indicators and our proposal fora synthetic indicator using the city of Barcelona, in Spain.

Principal component analysis (PCA) has been demonstratedto be a very useful technique with which to synthesize sets ofmonitoring indicators. PCA is a mathematical procedure thatuses a linear transformation to convert a set of correlatedvariables into a set of linearly uncorrelated variables namedprincipal components. The number of principal components isless than or equal to the number of original variables. Thistransformation is defined in such a way that the first prin-cipal component has the largest possible variance and eachsucceeding component in turn has the highest variancepossible. A reduction of data dimension can be obtained whilesubstantial information is retained in the new data set. Statedanother way, the ratio of information to the dimension of thenew data set is increased. In addition to the reduction of di-mension, PCA models can be utilized to construct two statisticsfrequently used in monitoring processes, the Q-statistic and theHotelling's T2 statistic. The Q-statistic corresponds to thenon-modeled, supposedly non-informative part of the data set.Therefore, this measure shows the lack of fitness of a (new)observation for the model. The Hotelling's T2-statistic is thesquared distance from the projection of a (new) observationonto the center of the model.

We can find some significant examples of the applicationof PCA in the urban context. Wong [5] uses PCA to examine therelationships amongst a set of indicators defining local economicdevelopment (LED). Based on a conceptual framework of 11factors widely perceived to be the major determinants of localeconomic development, 29 indicators were identified to mea-sure these factors. The author, besides examining the structure ofrelationships amongst the LED indicators compiled, used PCA toexplore the spatial patterns emerging from the analysis. A seriesof multiple regression models were then calibrated to investi-gate the relative strengths of relationships between the LEDindicators and suggested performance variables. Still in the areaof urban economics but turning to urban functional analysis,Chen et al. [6] used principal component analysis, cluster anal-ysis, and location quotient methods to analyze the economiesand industries of cities in the Xinjiang region of China, assessingfactors such as urban scale, growth pole level, specializationsectors and industry gradients. In the field of real estate econ-omics, Li [7] aimed to solve contradictions between the certaintyof fixed reference land prices at a certain time point and theuncertainty of land price changes,which lead to difficulties in theapplication of reference land prices using principal componentanalysis to select the main factors affecting land prices.

Focusing more on environmental monitoring, Liu et al. [8]utilized PCA methods to perform a qualitative and quantita-tive analysis of the spatial and temporal distribution of



3M.-L. Marsal-Llacuna et al. / Technological Forecasting & Social Change xxx (2014) xxx–xxx

various eco-environmental factors, in order to determine themain features that impact on the eco-environment. Tu andLin [9] conducted a principal component analysis using col-lected data sets of citizens' perceptions of residential envi-ronmental quality in order to extract the major scales andfactors of the residents' assessments. PCA was performedwith the results of a questionnaire consisting of 45 questionswhich was administered to 240 residents randomly select-ed from 80 residential buildings in residential–commercialmixed-use zones. Combining environmental issues with so-cial monitoring, Takano et al. [10] used PCA to study the asso-ciations between public spaces filled with greenery locatednear residential areas and easy to walk to, and the longevityof senior citizens in a densely populated developed megacity.

PCA can also be used to study urban patterns.Weng et al. [11]analyzed the spatial patterns of urban land surface temperatures(LST) and examined the factors contributing to LST variationsusing PCA, with potential factors being grouped into differentcategories. PCA helped determine the relative importance ofeach group of variables within its category. Qi et al. [12]constructed an index on urban land gradation which considered12 factors (commercial/services prosperity, road access, publicand external transport convenience, infrastructure, public ser-vices, cultural/sport facility maturity, natural conditions, envi-ronmental quality, population density, industry clustering andurban planning) to finally develop a system consisting of 35indicators. PCAhelped to analyze the index systemonurban landgradation. Ewing et al. [13] consolidated the variables whichdefine urban sprawl using PCA. The variables considered were

Fig. 1. List of strategies and their indicators contained in Bar


density, land use mix, degree of centering, and street accessibil-ity. These were then related to other factors such as vehicleownership, commutemode choice, commute time, vehiclemilestraveled per capita, traffic delay per capita, traffic fatalities percapita and 8-hour ozone level, to analyze how the factors affect-ed the variables and vice versa. The study of the correlations wasmade with multiple regression analysis. Finally, Li and Yeh [14]developed a PCA to reduce errors in multi-temporal images.Multi-temporal images are usually used to monitor urban ex-pansion for change detection, but tend to over-estimate land usechanges because classification signatures have been inadequate-ly created. The authors demonstrated that PCA could reduce sucherrors and provide a more reliable method for monitoring rapidland use changes and urban expansion.

Similar to Wong [5] we use PCA to analyze a set of in-dicators, but then go one step further by testing to aggregateindicators in a single index. As just seen in the state-of-the-art of PCA applications, there have been no attempts touse PCA to elaborate a value summarizing a data set. This isthe target of our experiment, as presented hereafter.

We use data corresponding to the Barcelona Local Agenda 21set of indicators for 11 years (from1999 to 2009) to obtain a PCAmodel. From the original set of 30 indicators described in Fig. 1,only 24 have been selected, because of lack of informationregarding the other 6.

Results from our experiment show how the set of indicatorsis extensively represented in the first component, accumulatingmore than 70% of variance. This means that all the originalindicators are highly correlated. For this reason, in our

celona's LA 21. Source: Ajuntament de Barcelona [4].



Fig. 2. Shows the loadings of the PCA model, representing how each indicator listed in Fig. 1 is represented in the construction of the LA 21 synthetic index.


experiment, only one principal component has been chosen asthe proposed LA 21 synthetic index. In a different experiment,if more than a principal component would be required, theHotelling's T2-statistic could be use to summarize the informa-tion in the proposed synthetic index.

Fig. 3. Shows the temporal evolution of the first PC of the PC


The way each original indicator is represented in the newindex can be easily observed by means of the so-called loadings.In our experiment (see Fig. 2), it is interesting to note that a sig-nificant number of indicators (14 of 24) have approximately thesame loading. This means that from a statistical point of view,

A model over 11 years for the indicators listed in Fig. 1.



Fig. 4. Table of values for Barcelona LA 21 indicators. This table shows the values for the set of indicators listed in Fig. 1 for the period 1999–2009.

3 In European countries, a large set of data can be obtained from the UrbanAudit of the European Office of Statistics (Eurostat). This monitors thequality of life in 321 cities with 338 indicators. The data set covers thefollowing domains: demography, social and economic factors, civic involve-ment, training and training provision, environment, travel and transport,information society, culture and recreation. Beside these objective data, theUrban Audit provides perception surveys as well. The last perception surveywas conducted in 2009 in 75 European cities. In random telephoneinterviews, 500 citizens per city were asked about their perception of thequality of life in their city (European Office of Statistics, Eurostat).


they are highly correlated. They have similar behavior during theobserved period of time, and consequently their role in the PCAmodel is also similar. Negative loadings show that the role ofsome indicators to the LA 21 synthetic index diverges from therest.

Fig. 3 shows the temporal evolution of the proposed syn-thetic index (first principal component) over 11 years (from1999 to 2009). In our experiment, the first PC successfullysummarizes the set of LA 21 indicators. The LA 21 summa-rizing index (Fig. 3) shows us how the city of Barcelona hasbetter compliance with LA 21 objectives every year. For theBarcelona case study, the utility of the proposed LA 21 indexis demonstrated.

Finally, the table below (Fig. 4) proves that PCA is a goodtechnique for representing the whole set of sustainabilityindicators (LA 21 indicators) in a synthetic index. We canidentify a positive increase in almost all of the values of theoverall set of indicators in the Barcelona LA 21 experiment.The summarizing indices obtained for each year (Fig. 3) showthe same positive increasing tendency.

2.2. Learning from the livable city and the successful QoL index

The livable city initiative monitors a city's welfare andwell-being with a quality of life (QoL) index. The term qualityof life should not be confused with the concept of standard ofliving, which is solely related to income. The standard set ofQoL indicators proposed by both Mercer and The EconomistIntelligence Unit includes data on wealth, employment,urban environment, social health, education, time-use, familyand community services. In both cases, an aggregated indexsummarizes a QoL set of indicators. QoL data is collected fromtwo different sources: life-satisfaction surveys of citizens,providing a subjective view of a population's emotional well-being in various life domains, and quality of life indicators, anobjective evaluation of more socio-economic factors. Results oflife-satisfaction surveys are primary data and hence surveys aredesigned specifically for that purpose. In contrast, objective


data to capture the social and economic dimensions comesfrom secondary sources such as a city's or country's statisticaloffices.3

Mercer publishes its quality of life survey annually. Initially, itwas conceived as a tool to help employees planning to workabroad and companies that were opening new businesses andneeded to know what conditions to offer their expatriateemployees. The survey measures 39 criteria grouped in 10 keycategories. The criteria have different weights according to theirimportance. New York serves as benchmark, with a score of 100,and other cities are ranked in relation to this baseline. The surveycompares 221 of the world's major cities [15].

The second widely recognized and important index measur-ing QoL is the one published by The Economist Intelligence Unit,which uses data from Mercer. However, unlike Mercer's TheEconomist index ranks both countries and cities. If we comparethe criteria used by the two organizations we quickly notice thatThe Economist uses fewer indicators: 9 as against 39. In reality,from a much wider set of indicators than the 9 displayed, TheEconomist selects the correlated variables which would explainmore than 80% of the inter-country variation using a regressionmodel, and ends up with the 9 shown below. To create the finalindex, these 9 indicators are weighted in an equation using theso-called Beta coefficients [16]. The indicators that weightedmost are health, material well-being, political stability andsecurity. These are followed by family relations and communitylife. Next and last in order of importance are, climate, jobsecurity, political freedom and gender equality [16].




Regarding indicators that were not considered in TheEconomist study (variables accounting for the remaining 20% ofinter-country variation), these were education level, rate of realGDP growth and income inequality (the Gini coefficient). Themost significant correlation amongst these variables was foundbetween education and overall life-satisfaction, as education hadan impact on incomes, health and possibly on political freedom[16].

Brief reference should be made to two international journalsthat publish lesswell-known annual lists ofmost livable cities. Intheir lifestyle magazine the media company Monocle (Monocle,top livable cities ranking) [17] lists the world's 25 top livablecities according to criteria similar to that seen in the previoussurveys. And, International LivingMagazine,which recommendsthe best places to retire to, travel to, buy real estate in and enjoylife, ranks 194 countries yearly according to their quality of life.To produce its index, the journal uses governmental sourcessuch as the World Health Organization, United Nations andothers like The Economist. Nine factors, similar to those seen forthe previously mentioned indices, are given weightings: cost ofliving, culture and leisure, economy, environment, freedom,health, infrastructure, safety and risk, and climate [18].

There is a growing interest in monitoring quality of lifenot only for the world's largest cities but also for others aswell, as a way to sell an image and attract new residents andbusinesses. Not being rated in the Mercer and The Economistlists is not an issue for small and medium sized cities; it isreasonable to suppose they can demonstrate elsewhere theirquality of life. Local and national governments are behindthese efforts. For instance, the UK Government has drawn uplocal quality of life indicator guidelines to support livabilitymonitoring in UK municipalities [19]. Accordingly, cities, Bristol[20] and even London (although ranked in the QoL surveysalreadymentioned) [21] publish their QoL scores in linewith theUK government's proposed indicators. Quality of life indicatorsare also drawn up at a regional level; South Gloucestershire [22]andWales [23] or Shropshire [24] county are examples.

In the US, we find QoL indicator guidelines published byuniversities, such as the University of Wisconsin [25]. As anexample of a city publishing the values of QoL indicatorswe haveJacksonville (Jacksonville Community Council) [26]. North WestIndiana is an example at a regional level [27].

In India, QoL indicator guidelines for cities have beenproduced by that country's government in collaboration withanNGO [28]. Those proposedby the government of NewZealand[29], or by the special administrative region ofHong-Kong [30] inChina, or the Canadian city of Edmonton [31] are further exam-ples of governments working to disseminate the quality of life oftheir cities.

The research community has also shown itself to beinterested in the subject of quality of life, for the most partfocusing on the study of the monitoring indicators. In generalterms, Senecal and Hamel [32] relate urban quality of life anda city's sustainability with a compact model pattern. Still in ageneral frame, Fahy and Cinneide [33] have researched QoLcriteria to produce an operational framework to assess quality oflife in an urban setting. The same authors [34], in amore focusedstudy, state that subjective QoL indicators are unlikely to beacceptable or particularly useful unless they are designed in closeconsultationwith the target populations. The authors defend theuse of community derived quality of life indicators instead of


standard life-satisfaction surveys. Campanera and Higgins [35]have researched QoL indicators that would allow better com-parison between urban-classified and rural-classifiedmunicipal-ities. Lee [36] proposes a set of indicators to evaluate the qualityof life of urban dwellers with regard to the semi-public spaces ofhigh-rise mixed use housing (HMUH) complexes in order tohelp in the design ofmore livable complexes. Lastly, Galster et al.[37] have studied the key factors eroding a neighborhood qualityof life. This study was carried out in US metropolitan areas. Thefindings showed that the poverty rate, male unemployment, theoverall adult unemployment rate, the female headship ratefor families with children, and the secondary school drop outrate were the most important factors in determining aneighborhood's quality of life.

Although the efforts of both governments and academiato research and improve QoL indicators for small andmedium-sized cities are of great value, a synthetic index isessential to make these efforts both more comparable andeasier to assess. While Mercer's and The Economist Intelli-gence Unit elaborate QoL indices on a yearly basis, and thejournals Monocle and International Living Magazine rankcities annually on the basis of livability, initiatives in the areaof quality of life and livability assessment led by govern-ments and academia are intermittent and do not generatemuch interest. This is because both governments and academiaare focused on content, elaborating a very comprehensive list ofindicators to assess quality of life and livability, but not asummary indexwhichwould allowcomparability. QoL indices ofMercer's and The Economist Intelligence Unitmake it possible tocompare cities and countries, and thereby establish rankings.Monocle and International Living Magazine directly elaboraterankings, based on the indices they develop to summarizelivability indicators.

Mercer's and The Economist Intelligence Unit QoL indicesare used as benchmarks in many contexts. Monocle andInternational Living Magazine rankings also serve as a globalreference in terms of livability. This is because of the com-parability these indices allow. Therefore, it is possible to saythat these non-official indices somehow normalize the terrainsof QoL and livability. In order to design indices with widerperspective and representation, we would suggest that officialglobal and country standardization organizations start elabo-rating summarizing indices. These official summarizing indicesshould be the basis of any ruled normalization process. In thatregard, in the last section of this paper, we provide some adviceto normalization bodies that already have started to standard-ize the Smart Cities initiative.

3. Contributing to the monitoring of the SmartCities initiative

3.1. The advantages of a summarizing index

We have found two Smart Cities monitoring initiativesalready in operation, but both producing indicators fromhistorical statistical data.

The first initiative in the field of Smart Cities monitoringcomes from academia. A consortium made up of the ViennaUniversity of Technology, University of Ljubljana and DelftUniversity of Technology [38] has produced a ranking ofEuropean cities based on an index of “smartness” consisting




of 74 indicators in six categories. This academic initiativewould be a breakthrough in the emerging area of themonitoringof cities smartness, if the proposed index did not result forma setof indicators based onhistorical data. The ranking is composed of70 medium-sized cities with urban populations of between100,000 and 500,000 inhabitantswith at least one university anda catchment area of less than 1,500,000 inhabitants (to excludecities dominated by a bigger city) [38]. The indicators werecreated using the Urban Audit database, produced by theEuropean Statistical Office (Eurostat) [39] and organized in thefollowing categories: Smart Economy, Smart Mobility, SmartGovernance, Smart Living, Smart People and Smart Environ-ment. Examples of indicators for the Smart Living categoryare: number of cultural facilities, health conditions, individ-ual safety, housing quality, number of educational facilities,touristic attraction and social cohesion. To compare thedifferent indicators values are standardized. The methodutilized is the z-transformation in which all indicator values aretransformed into standardized values with a 0 average and astandard deviation of 1. To obtain results at the category level,indicators must be aggregated. In the aggregation, the represen-tation of indicators is considered, weighting indicators accordingto coverage. In this context, we have to understand representa-tion of indicators as the coverage of the sample represented bythe indicator. Therefore, for implementations of the describedmethod, indicators will be weighted according to the number ofsegment(s) of represented population. The aggregation isadditive, meaning that totals are divided by the number ofvalues added. No further corrections or weightings are made toproduce a city's final index of “smartness”.

Still in academia, before presenting the second initiative,it is worth mentioning the studies conducted by the researchcommunity and their findings concerning monitoring of theSmart Cities initiative. Cargaliu et al. [40] have also used theEU Urban Audit data set to analyze the factors determining theperformance of Smart Cities. They found that the presence of acreative class, attention to the urban environment, level ofeducation, multimodal accessibility and the use of ICTs for publicadministration are all positively correlated with urban smart-ness. More interested in summarizing indices thanworkingwitha set of indicators, Malek [41] studied the suitability of anexisting index, the Informative Global Community DevelopmentIndex (IGC), for monitoring the Smart Cities initiative.

The second is the European Initiative on Smart Cities,promoted by the European Commission, or more specificallythe Strategic Energy Technologies Information System (SETIS)[42]. This initiative focuses on achieving more efficient modelsfor reducing carbon emissions in Europe. Its specific objectivesare to involve aminimumof 5% of the European population inthe use of low carbon emission technologies as well as toreduce by 40% greenhouse gas emissions by 2020 and tospread across Europe best practices in the area of sustain-able energy. The initiative proposes specific actions forbuildings, energy networks (heating, cooling and electric-ity) and transport, and seeks the participation of 25 largecities (N500,000 inhabitants) and five very large cities(N1,000,000 inhabitants) to commit to implementingprograms in these three sectors. A set of key performanceindicators (KPI) is proposed. For energy networks, theseinclude: meeting 50% of heat and cooling demand from re-newable energy sources (RES), establishing at least 20 pilot


schemes by2015 for “smart grids” coupledwith “smart building”equipment, and measuring energy consumption with “smartmeters” [42]. Thismonitoring initiative has a clear focus on smartenergy and differs from the previous one in that it does not havea summarizing index, but sets of key performance indicators tomeet, which is less desirable for monitoring. In this initiative,what needs to be highlighted is that every participating citybuilds their indicators on the basis of energy consumption anddemand, making these indicators more intelligent.

Having studied the need for summarizing indices to monitorthe Smart Cities initiative, and before considering the need forintelligence in the proposed index, it is important tomention theclose connection between the quality of life concept and the“transitional initiatives” to the Smart Cities concept. RichardFlorida's concept of the creative city (2002) is a compellingargument about urban development and its dependency onnovel combinations of knowledge and ideas. The concept isbased on the assertion that certain occupations specialize in thecreative task. People in these occupations are drawn to areasproviding a high quality of life, and therefore it should be anessential development strategy for cities to create an environ-ment that attracts and retains thoseworkers [43]. More or less inparallel with the concept of the creative city appeared theconcept of the digital city (2004) or e-city, focusing more on theidea of a “connected community”, combining broadbandcommunications with government open data and open industrystandards as a way to enhance the quality of life of “e-citizens”[44]. Before going further into the smart city concept we shouldmention the knowledge city concept (2008), with closerconnections to specialized education as a key issue for ensuringa society's long-term quality of life. The Knowledge City can beconsidered a “bridging initiative” to the Intelligent City [45] orbetter known Smart City.

Despite its apparent novelty, the concept of Smart Cities wascoined back in the early nineties. In 1993, Gibson, Kozmetskyand Smilor, in their book The Technopolis Phenomenon — smartcities, fast systems, global networks [46] anticipate a kind ofurban-tech phenomenon to come that, through a twenty-firstcentury infrastructure, would contribute to the enhancement ofthe quality of life as well as widening the range of globalmarketplaces. The authors' vision was that academia, togetherwith governments and industry, would present information,ideas, programs and initiatives in a new manner, much moretechnological and informed than previously, that would accel-erate the creation of smart cities, fast systems, and global networks.They were not that wrong. In 1999 we find the earliest appli-cation of Smart Cities experiences. Mahizhnan [47] presents thecase of Singapore, whose transformation into an informationeconomy on the back of information technologies was madenecessary by the island's lack of natural resources plus the needto reinvent its traditional industrial economy. Nearly contempo-raneously, in 2001, we find the experience of Edinburgh, wherethe government made a huge investment in the necessary tech-nological infrastructure to turn the city into an experimental ITcenter, and where the first e-Government initiatives took place.

But despite its earlier development, the Smart Cities conceptonly becamewidely known after 2009, and even today it is still asomewhat fuzzy idea. In an attempt to provide an overalldefinition of the concept we could say that Smart Cities haveevolved out of livable, creative, digital and knowledge cities,drawing heavily on the concept of the Sustainable City and




having in common a large technological component. A less con-ceptual definitionwould be that the Smart Cities initiative tries toimprove urban performance by using data, information andinformation technologies (IT) to provide more efficient servicesto citizens, to monitor and optimize existing infrastructure, toincrease collaboration amongst different economic actors and toencourage innovative business models in both the private andpublic sectors. Adding to both definitions, in a more technicaldescription, we could say that cities wishing to become smartmust be equippedwith a “brain” (software) suppliedwith lots ofreal time information (data collected from sensors) enablingthem to take more sustainable, efficient and citizen-centricdecisions, smoothly transforming decisions into actions bymeans of technological solutions.

On the basis of these definitions and technical descriptionswe can anticipate an important issue with regard tomonitoring:Smart Cities initiative must be monitored as close as possible toreal time because it seeks to measure cities' urban performanceon the basis of high IT levels, it cannot be monitored withstatistics based on historical data.

The Smart Cities concept's point of departure is thatindividuals are permanently connected with others and withtheir machines. This hyper-connection relates to the commonimage of users connecting not only with their smart phone butalso with their washing machine, fridge, heating, lights, etc.through sensors in connection with their smart phones. Intechnical terms, the connecting network between digital devicesand sensors is given the name Internet of Things (IoT). Eachdevice is equipped with sensors so that humans and othermechanisms can operate them at a distance. Here, it is worthmentioning that in today's cities, people, companies, govern-ments etc. generate digital data on almost all the urban activitiesthey perform, but it is estimated that only 5% of the availabledigital information is currently being used. This hyper-connection in society anticipates the idea of the “sensor-citizen”.

In the Smart Cities context all these technological intercon-nections are known as ambient intelligence. Several studies havedeveloped standardized frameworks for its application to SmartCities [48]. A similar concept is ubiquitous computing. The ex-ample of South Korean Smart Cities promotes the developmentof standard architecture to integrate both ubiquitous computingand ambient intelligence in green technologies [49]. The tech-nological possibilities offered by ambient intelligence, togetherwith ubiquitous connectivity, the fast Internet and a dramaticcost reduction in sensors, suggest extremely well monitoredcities in the near future.

If we consider these extremely well monitored cities of thefuture and today's fast-evolving society, which is changingmorerapidly than statistics are being updated, we are forced to theconclusion that we will have to use sources other than historicalstatistics to monitor the Smart Cities initiative. In our opinion,indicators based on historical statistical information should besubstituted by indicators constructed out of real time data. SmartCities need “smart indicators”: information currently extractedfrom the city, which duly processed is converted into indicatorsand returned to the city as monitoring values.

3.2. Summarizing indices require real-time data

We have already stated our view that the monitoring ofSmart Cities needs a set of smart indicators summarized in a


synthetic index. These smart indicators would be obtainedfrom data in as real time as possible. A set of smart indicatorswould monitor a city's technological and information ad-vances, quality of life and sustainability improvements. Tomonitor all these aspects of the Smart City, real time physicaland non-physical data would be needed.

In the context of our proposal, about the need to obtain datain real time to elaborate smart indicators and indices, we do nothave to understand “real time” in terms of computing. We usethe concept of real time to reflect what is happening “now” or“live”, rather than to measure and update data every second.

Today's technology providing the most updated data oncities' physical recognition is remote sensing imagery. Itwould take too long to mention just some relevant examplesof urban pattern recognition for the physical monitoring ofcities thus there are several successful experiences in thatarea of application. In the urban context, remote sensingimagery has been used not only for pattern recognition butalso for building features extraction and automatic recon-struction; here, we will also avoid having to draw up a longlist by not referring to examples. Instead, we will mentionsome techniques based on remote sensing data that makepossible a more detailed study of urban pattern changes. Themost traditional methodologies for analyzing settlement pat-terns on the basis of remote sensing imagery are principal com-ponent analysis (to determine differentiation patterns), clusteranalysis (to determine homogeneity patterns), and regressionanalysis (to identify the explicit functional relations betweensettlement patterns and their underlying variables). Additional-ly, Rebelo [50] has studied the use of decision trees, neural net-works and the link between neural networks and the nodes ofdecision trees for urban settlements pattern classification. In thefield of simulation, Han et al. [51] aim to predict urban growth byadding cellular automata techniques to satellite imagery. Alsousing simulation techniques, Schwarz et al. [52] and Schetcke[53] have studied the opposite phenomena: urban shrinkage.

When we focus on the more common and accessibleapplications of satellite imagerywe find, in 2006, the appearanceof Google Maps. In combination with the GIS tools that alreadyexisted, these changed the way location aware systems worked[54]. Internet mapping, despite being a big step forward becauseit sets aside traditional cartographic techniques depending onexpensive specific photographic flights, is not and cannot showreal time data. Thus, satellite-imagery-based maps are onlyperiodically updated. What can be and already is in real time isthe positioning of the objects in the city, on the basis of Internetmapping. Today, using location aware systems, traffic andmobility in the city can be easily monitored in real time. Acomputerized system is constantly receiving changing data andis processing it sufficiently rapidly to be able to monitor thesource of the data, enabling the system to direct or actuate themonitored process, if needed.

Besides the physical urban data, the non-physical oneobtained from the city needs to be monitored. Non-physicaldata refers to information on the provisioning and consump-tion of urban services, from energy to water supply to wastecollection. Non-physical urban data is information on supplyand demand, and therefore it is much more challenging tomonitor than physical data. For the non-physical urban data,data collection depends on the collaboration of utility companiesand the participation of the citizens (citizen-sensor).




To provide an example of what could be smart indicators inthe area of non-physical urban data and show their advantagesover parallel information obtained from historical statisticalsources, we have studied the occupancy and utilization of builtstocks in two Catalan cities over a 5-year period. Statistical datawas extracted from the Spanish National Institute of Statistics(Instituto Nacional de Estadística) [55] and data on energyconsumptions was kindly provided by the Spanish energycompany, FECSA-ENDESA Energia 21. Data provided by theenergy companywas partially obtained from the so-called smartgrids. Smart grids have more powerful monitoring applicationsthan that of merely providing information about total amountsof energy consumption; they can even monitor the contributionof renewables to the net. Smart grids exemplify the concept ofproducers and energy consumers being in constant communi-cation in order to achieve amore efficient, profitable and reliablesupply of energy. It is clear, then, that indicators for monitoringsmart energy (e.g.) will not be the number of solar panelsinstalled, for example, but their real time contribution to theenergy net.

Remaining in the area of smart energy, a further exampleon how smart indicators could be obtained would be thedata processed out of so called smart meters. Here, the citizencollaborates in obtaining data. For instance, The GooglePowerMeter visualizes individual energy usage by capturingdata fromsmartmeters. Google.org, promoters of the initiative–currently solely implemented in the US – state that it wasdesigned as a free monitoring tool to raise awareness about theimportance of giving people access to their energy information.Besides visualization, the tool includes the feature of sharinginformation with others through smart phones and the

Fig. 5. Statistics of built park and energy consumption


possibility of receiving personalized recommendations onsaving energy. On June 24, 2011, Google announced thatthey were discontinuing the PowerMeter because manufac-turers and their utilities partners were opening direct accessto their energy information services.

Back to our experiment, the figures below compare thelatest statistical data available (Spanish Census, 2001) [55] onthe occupation of residential stocks (main residence, holidayresidence or unoccupied residential units), the utilization ofcommercial buildings (with and without activity) and in-dustrial installations (differentiating between services and fa-cilities, commercial and industrial), with the number ofelectrical supply contracts in the same categories. Billedenergy and billed power are also shown in the figures.

More than a simplemonitoring of energy consumption usingcollected data, the updated information provided by the elec-tricity company can help in the monitoring of more hidden orlatent phenomenon affecting cities such as the current economiccrisis. We can observe how the number of general clients'contracts (domestic clients and small businesses up to 15 kW ofsupplied power) and consumption started to decrease in 2008 inboth cities (Fig. 4). But, interestingly, we can observe the op-posite with medium sized businesses and large companies: thenumber of clients, billed energy and power rose considerably(Fig. 5). This suggests that the current crisis has hit families andsmall businesses particularly hard (Fig. 6).

This experiment shows that non-physical real time data isvaluable for more than the immediate purpose of monitoringenergy consumption. Initially, the monitoring of real timedata on consumption makes it possible to model consumerbehavior, and design demand policies that avoid peaks and

measurements: aggregated and residential data.



Fig. 6. Statistics of built park and energy consumption measurements: industrial and business sectors data.


valleys and incentivize or reduce demand in certain areas,etc. Additionally, through actuators, an energy system can beforced to collaborate with a renewable system when needed.Interestingly, besides this immediate monitoring result, ourexperiment revealed that further monitoring can be extract-ed. Indeed, real time data on energy consumption reveals thereal usage of the city, which is extremely useful for social andeconomical policy making, since more accurate design andgreater acceptance of policies is likely, based on updatedinformation.

4. Conclusions and further research: the need for an“intelligent index” to monitor the Smart Cities initiative

These conclusions recap the lessons of previous experiencesof city monitoring, the proposals that came out of what waslearnt and their specific application to the Smart Cities initiative.

Firstly, the need for an aggregated or synthetic index tovisualize monitoring results from a set of indicators has beenconclusively demonstrated. Indices, apart frommaking compar-isons and rank order lists possible, summarize large amountsof data. In an era of information excess, where informationfilters and synthesis is essential, an index summarizing a city'ssmartnesswould help guarantee an initiative's success. Differenttechniques for obtaining summarizing indices have been shown.We would specifically propose the use of principal componentanalysis because its feasibility has been demonstrated by theexample of Barcelona's LA 21. The experiences of private bodiesand academic research have shown how mathematical modelsare also a promising technique.

Regarding the proposal for a synthetic index, we advisedthat official normalizing bodies have to take the lead in theelaboration of summarizing indices, as the basis of processes tonormalize urban initiatives. In that regard, this advice is usefulfor the on-going normalization of the Smart Cities initiative,and should reach organizations such as ISO, CEN-CENELEC,


AENOR (Spanish Normalization Organization) and othercountry normalization organizations. All of these organizationshave started normalization processes at international, Europe-an, and national levels, to normalize the Smart Cities initiative.

The International Standards Organization (ISO) has initiatedstandardization activities in the field of Smart Cities. ISO stan-dardizations on Smart Cities are the first initiative of this kind.ISO standardization objectives are to obtain certification rules.Certification rules are non-binding, meaning that countries,industry, and stakeholders are free to adopt the recommenda-tions and standards provided by the rules. However, as ISOstandardization rules are going to be the first of this kind, theywill pave theway for further standardization, either at Europeanor national levels.

The European Committee for Standardization (CEN),depending on the EU Commission, is about to start standard-ization activities at the European level. By EU mandate, CENproduces normalization rules which differ from certificationbecause they are binding. Therefore, when CEN normalizationrules for Smart Cities see the light, they will be mandatory for allMember States. CEN rules for the normalization of Smart Cities inEurope build on ISO certification works.

Additionally, some European countries have started theirrespective standardization activities at the national level,which normally consists of normalization works (e.g. AENOR,in Spain; AFNOR, in France). On-going EU National standard-ization works build on ISO, since CEN is just starting itsstandardization activities.

In 2012 ISO created the ISO/TC 268 Committee on “sustain-able development in communities”. The Committee has twoworking groups: ISO/TC 268/WG 01 “Management Systemsstandards for sustainable development in communities” andISO/TC 268/WG 02 “Urban indicators”, which are responsible ofthe following projects: Project ISO 37101 Internal DesignSpecifications, Project ISO 37120 Sustainable Development andResilience. Indicators for City Services and Quality of Life.




ISO standardizationworks are at the stage of State-of-the-Art(SotA) in the two focus areas (corresponding to workinggroups): smart city management and smart city indicators. Inthe management field, the latest documents list general prin-ciples and requirements formanaging Smart Cities, and verifyingcompatibility and consistency with existing related rules. In thearea of smart city indicators, more advanced works collectexisting indicators on Sustainable Development andResilience inCities (TR 37121) and Global City Indicators for City Services andQuality of Life.

The Committee has created the subcommittee: ISO/TC 268/SC 01 “smart urban infrastructure metrics”, which in turn hascreated a working group, ISO/TC 268/SC 01/WG 01 “infrastruc-ture metrics”, responsible for the Project TR37150 “Review ofworks relevant to smart community infrastructure metrics andfuture directions of standardization”.

In the ISO on-going production of documents we can alreadyfind sets of proposed indicators to measure different aspects of aSmart City, but no proposal for summarizing indices. We hopesummarizing indiceswill appear in the next set of documents. Asone of the outcomes of this paper, we strongly encouragenormalization bodies to start elaborating indices to summarizesets of related indicators, to effectively monitor and easily com-pare the smartness of our cities.

Secondly, in this paper, we stressed the need for real-timeindicators (what we call smart indicators) as the appropriatemonitoring tool for developing a final smart or intelligent index.It has been illustrated how historical information does not allowtaking immediate reaction as its frequency update is not highenough. The Smart Cities initiative attempts to asses what ishappening in the present in a city in terms of urban IT, quality oflife and sustainability. Additionally, in one of the case studiespresented in this paper, through the analysis of energy con-sumption, we learned that even monitoring the quality of lifeand sustainability (issues apparently static and well covered byhistorical statistics) can be improved with the use of real timedata, allowing better andmore accurate design of social policies.Therefore, there is no question about the need for real time datato illustrate the smartness of a city.

The basic concept of the Smart Cities initiative can beexpressed as follows: the Smart Cities initiative seeks to improveurban performance by using data, information and InformationTechnologies (IT) to provide more efficient services to citizens, tomonitor and optimize existing infrastructure, to increase collabora-tion between different economic actors and to encourage innovativebusiness models in both the private and public sectors. There havealready been examples of successful collaboration betweendifferent actors – citizens, private and public bodies – in thisrespect, and we have shown these. We strongly believe that theamount of real time data needed to create smart indicators tofurther develop smart indices will continue growing, and thatthese are the first two necessary steps towards guaranteeing thesuccess of the Smart Cities initiative.

References

[1] J. Colding, ‘Ecological land-use complementation’ for building resil-ience in urban ecosystems, Landsc. Urban Plan. 81 (1–2) (2007) 46–55.

[2] J.R. Barton, Urban sustainability as strategic planning, EURE-Rev.Latinoam. Estud. urbano regionales 32 (96) (2006) 27–45.

[3] R. Tuts, Localizing Agenda 21 in small cities in Kenya, Morocco andVietnam, Environ. Urban. 10 (2) (1998) 175–189.


[4] Ajuntament de Barcelona, Indicadors 21. Indicadors locals de sostenibilitat aBarcelona, 2003. (Documents 8).

[5] C. Wong, Developing indicators to inform local economic developmentin England, Urban Stud. 39 (10) (2002) 1833–1863.

[6] C.P. Chen, Y.J. Ding, S.Y. Liu, City economical function and industrialdevelopment: case study along the railway line in North Xinjiang inChina, J. Urban Plan. Dev.-ASCE 134 (4) (2008) 153–158.

[7] S. Li, Multi-factor prediction of urban reference land price based onprincipal component analysis and BP network, Proceedings of the 5thAsia-Pacific Conference on Control & Measurement, 2002, pp. 269–273.

[8] H.Q. Liu, H.L. Gong, Z.W. Hu, Ecological environment regionalization ofBeijing: RS & GIS application research, 2009, Joint Urban Remote Sens.Event 1–3 (2009) 190–196.

[9] K.J. Tu, L.T. Lin, Evaluative structure of perceived residential environ-ment quality in high-density and mixed-use urban settings: anexploratory study on Taipei City, Landsc. Urban Plan. 87 (3) (2008)157–171.

[10] T. Takano, K. Nakamura, M. Watanabe, Urban residential environmentsand senior citizens' longevity in megacity areas: the importance ofwalkable green spaces, J. Epidemiol. Community Health 56 (12) (2002)913–918.

[11] Q.H. Weng, H. Liu, B.Q. Liang, D.S. Lu, The spatial variations of urban landsurface temperatures: pertinent factors, zoning effect, and seasonalvariability, IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 1 (2) (2008)154–166.

[12] X.Z. Qi, S.Z. Zhao, W. Xu, W.Y. Jiang, Research on the index system of urbanland gradation based on principal component analysis, in: J.P. Xu, Y. Jiang, Y.Tan (Eds.), Proceedings of the Second International Conference onManagement Science and Engineering Management, Management Scienceand Engineering Management, 2008, pp. 555–563.

[13] R. Ewing, R. Pendall, D. Chen, Measuring sprawl and its transportationimpacts, travel demand and land use 2003: planning and administra-tion, Transportation Research Record, 1831, 2003, pp. 175–183.

[14] X. Li, A.G. Yeh, Principal component analysis of stacked multi-temporalimages for the monitoring of rapid urban expansion in the Pearl RiverDelta, Int. J. Remote Sens. 19 (8) (1998) 1501–1518.

[15] Mercer's Quality of Living Survey, http://www.mercer.com/articles/quality-of-living-definition-1436405.

[16] The Economist Intelligence Unit, Quality of Life Survey, http://www.economist.com/media/pdf/QUALITY_OF_LIFE.pdf.

[17] Monocle, Top Livable Cities ranking, http://www.monocle.com/sections/edits/Web-Articles/Top-25-Cities/.

[18] International Living, Cities' quality of life ranking, http://internationalliving.com/2010/02/quality-of-life-2010/.

[19] UK Audit Commission, http://www.audit-commission.gov.uk/nationalstudies/firerescue/Pages/localqualityoflifeindicators_copy.aspx.

[20] Bristol Municipality, http://www.bristol.gov.uk/node/4103.[21] LondonMuncipality, http://www.londonsdc.org/sustainable_development/

monitoring.aspx.[22] SouthGloucestershire Council, http://www.southglos.gov.uk/Communities/

QualityofLifeIndicators.htm.[23] Wales Council, http://www.dataunitwales.gov.uk/Data.asp?cat=285.[24] Shropshire County, http://www.shropshire.gov.uk/factsfigures.nsf/

open/3B9A853832652F2D80256FB200331E32.[25] University of Wisconsin, http://www.uwex.edu/ces/cced/communities/

qualityoflifedataindicatorsdatasources.cfm.[26] Jacksonville Community Council, http://www.jcci.org/jcciwebsite/pages/

indicators.html.[27] Northwest Indiana Quality of Life Council, http://www.nwiqlc.org/.[28] Indian Government and Janaagraha, http://www.janaagraha.org/

civicissue/quality-life-indicators.[29] New Zealand Government, http://www.bigcities.govt.nz/.[30] Hong-Kong Muncipality, http://www2.cuhk.edu.hk/ssc/qol/eng/hkqol.

html.[31] Edmonton Municipality, http://www.edmonton.ca/city_government/

planning_development/neighbourhood-indicators.aspx.[32] G. Senecal, P.J. Hamel, Compact city and quality of life: discussions

of the Canadian approach to sustainability indicators, Can. Geogr.(Geographe Canadien), 45 (2) (2001) 306–318.

[33] F. Fahy, M.O. Cinneide, Community-based quality of life indicators forurban areas as derived in Galway City, Ireland, in: U. Mander, C.A.Brebbia, E. Tiezzi (Eds.), Sustainable City IV: Urban Regeneration andSustainability, WIT Transactions on Ecology and the Environment,2006, pp. 691–702.

[34] F. Fahy, M.O. Cinneide, Developing and testing an operational frameworkfor assessing quality of life, Environ. Impact Assess. Rev. 28 (6) (2008)366–379.

[35] J.M. Campanera, P. Higgins, Quality of life in urban-classified andrural-classified English local authority areas, Environ. Plan. A 43 (3)(2011) 683–702.


http://refhub.elsevier.com/S0040-1625(14)00045-6/rf0005










































http://www.mercer.com/articles/quality-of-living-definition-1436405

http://www.mercer.com/articles/quality-of-living-definition-1436405

http://www.economist.com/media/pdf/QUALITY_OF_LIFE.pdf

http://www.economist.com/media/pdf/QUALITY_OF_LIFE.pdf

http://www.monocle.com/sections/edits/Web-Articles/Top-25-Cities/

http://www.monocle.com/sections/edits/Web-Articles/Top-25-Cities/

http://internationalliving.com/2010/02/quality-of-life-2010/

http://internationalliving.com/2010/02/quality-of-life-2010/

http://www.audit-commission.gov.uk/nationalstudies/firerescue/Pages/localqualityoflifeindicators_copy.aspx

http://www.audit-commission.gov.uk/nationalstudies/firerescue/Pages/localqualityoflifeindicators_copy.aspx

http://www.bristol.gov.uk/node/4103

http://www.londonsdc.org/sustainable_development/monitoring.aspx

http://www.londonsdc.org/sustainable_development/monitoring.aspx

http://www.southglos.gov.uk/Communities/QualityofLifeIndicators.htm

http://www.southglos.gov.uk/Communities/QualityofLifeIndicators.htm

http://www.dataunitwales.gov.uk/Data.asp?cat=285

http://www.shropshire.gov.uk/factsfigures.nsf/open/3B9A853832652F2D80256FB200331E32

http://www.shropshire.gov.uk/factsfigures.nsf/open/3B9A853832652F2D80256FB200331E32

http://www.uwex.edu/ces/cced/communities/qualityoflifedataindicatorsdatasources.cfm

http://www.uwex.edu/ces/cced/communities/qualityoflifedataindicatorsdatasources.cfm

http://www.jcci.org/jcciwebsite/pages/indicators.html

http://www.jcci.org/jcciwebsite/pages/indicators.html

http://www.nwiqlc.org/

http://www.janaagraha.org/civicissue/quality-life-indicators

http://www.janaagraha.org/civicissue/quality-life-indicators

http://www.bigcities.govt.nz/

http://www2.cuhk.edu.hk/ssc/qol/eng/hkqol.html

http://www2.cuhk.edu.hk/ssc/qol/eng/hkqol.html

http://www.edmonton.ca/city_government/planning_development/neighbourhood-indicators.aspx

http://www.edmonton.ca/city_government/planning_development/neighbourhood-indicators.aspx

















[36] J. Lee, Quality of life and semipublic spaces in high-rise mixed-usehousing complexes in South Korea, J. Asian Archit. Build. Eng. 10 (1)(2011) 149–156.

[37] G. Galster, R. Quercia, A. Cortes, Identifying sustainable neighborhoods: anempirical exploration of threshold effects, in: C.A. Brebbia, A. Ferrante, M.Rodriguez, B. Terra (Eds.), Sustainable City: Urban Regeneration andSustainability, Advances in Architecture Series, 9, 2000, pp. 371–380.

[38] Vienna University of Technology, University of Ljubljana, Delft University ofTechnology, European Smart Cities ranking, http://www.smart-cities.eu/.

[39] European Office of Statistics (Eurostat), Urban Audit database, http://epp.eurostat.ec.europa.eu/portal/page/portal/region_cities/city_urban.

[40] A. Caragliu, C. del Bo, P. Nijkamp, Smart cities in Europe, CERS 2009 —

3rd Central European Conference in Regional Science, InternationalConference Proceedings, 2009, pp. 45–59.

[41] J.A. Malek, Informative global community development index of informa-tive Smart City, in: R. Revetria, V. Mladenov, N. Mastorakis (Eds.), EDU '09:Proceedings of the 8th WSEAS International Conference on Education andEducational Technology, Recent Advances in Computer Engineering, 2009,pp. 121–125.

[42] Strategic Energy Technologies Information System (SETIS), EuropeanInitiative on Smart Cities, http://setis.ec.europa.eu/about-setis/technology-roadmap/european-initiative-on-smart-cities.

[43] D. McGranahan, T. Wojan, Recasting the creative class to examine growthprocesses in rural and urban counties, Reg. Stud. 41 (2) (2007) 197–216.

[44] M. Craglia, L. Leontidou, G. Nuvolati, J. Schweikart, Towards the develop-ment of quality of life indicators in the ‘digital’ city, Environ. Plan. B-Plan.Des. 31 (1) (2004) 51–64.

[45] K. Rodrigues, E. Tome, Knowledge cities: a Portuguese case, Proceedings ofthe 3rd European Conference on Intellectual Capital, 2011, pp. 350–358.

[46] P. Harris, The technopolis phenomenon — smart cities, fast systems,global networks, Behav. Sci. 38 (2) (1993) 141–143.

[47] A. Mahizhnan, Smart cities— the Singapore case, Cities 16 (1) (1999) 13–18.[48] N. Streitz, Ambient intelligence research landscapes: introduction and

overview, ambient intelligence, in: B. DeRuyter, R. Wichert, D.V. Keyson,P. Markopoulos, N. Streitz, M. Divitini, N. Georgantas, A.M. Gomez (Eds.),Lecture Notes in Computer Science, 6439, 2010, pp. 300–303.

[49] J. Lee, S. Baik, C. Lee, Building an integrated service management platformfor ubiquitous ecological cities, Computer 44 (6) (2011) 56–63.

[50] E.M. Rebelo, A methodology to approach immigrants' land use inmetropolitan areas, Cities 27 (3) (2010) 137–153.


[51] J. Han, Y. Hayashi, X. Cao, H. Imura, Application of an integrated systemdynamics and cellular automata model for urban growth assessment: acase study of Shanghai, China, Landsc. Urban Plan. 91 (3) (2009) 133–141.

[52] N. Schwarz, D. Haase, R. Seppelt, Omnipresent sprawl? A review of urbansimulation models with respect to urban shrinkage, Environment andPlanning B-Planning & Design 37 (2) (2010) 265–283.

[53] S. Schetke, D. Haase, Multi-criteria assessment of socio-environmentalaspects in shrinking cities. Experiences from eastern Germany, Environ.Impact Assess. Rev. 28 (7) (2008) 483–503.

[54] S.C. Tan, O. Wong, Location aware applications for smart cities with Googlemaps and GIS tools, in: Y. Li, M. Looi, N. Zhong (Eds.), Frontiers in ArtificialIntelligence and Applications, Advances in Intelligent IT: Active MediaTechnology 2006, 138, 2006, pp. 223–228.

[55] Instituto Nacional de Estadística (National Statistics Institute), Spainhttp://www.ine.es/.

Maria-Lluïsa Marsal-Llacuna, Architect (2001), PhD in Urban Planning(2008), both at Universitat Politécnica de Catalunya (UPC). MSc in IndustrialInformatics, Automation and Systems Computation (2012) and PhD inTechnology (2013), both at Universitat de Girona (UdG). Associate professorat the Spatial and Urban Planning Department (UPC) from 2001 until 2009.Professor at the Department of Architecture, Urban Planning Area (UdG)since 2009. Member of eXiT, Research Group on Control Engineering andIntelligent Systems.

Joaquim Meléndez-Frigola received the B.S. degree in Telecom Engineeringfrom the Universitat Politècnica de Catalunya in 1991 and his PhD from theUniversitat de Girona in 1998. Professor at the Department of Electrical,Electronic and Automation Engineering of the Universitat de Girona (Spain)since 1999 and Director of the Phd Program in Technology at the sameuniversity. Director of eXiT, Research Group on Control Engineering andIntelligent Systems.

Joan Colomer-Llinàs, Telecom Engineering from the Universitat Politècnica deCatalunya and PhD from the Universitat de Girona. Professor, Secretary (2001)and Director (2006) at the Department of Electrical, Electronic and AutomationEngineering of the Universitat de Girona (Spain) Co-director of eXiT, ResearchGroup on Control Engineering and Intelligent Systems.









http://www.smart-cities.eu/

http://epp.eurostat.ec.europa.eu/portal/page/portal/region_cities/city_urban

http://epp.eurostat.ec.europa.eu/portal/page/portal/region_cities/city_urban









http://setis.ec.europa.eu/about-setis/technology-roadmap/european-initiative-on-smart-cities

http://setis.ec.europa.eu/about-setis/technology-roadmap/european-initiative-on-smart-cities
































http://www.ine.es/
