ERCIM News 98

Also in this issue:

Keynote:

Smart Cities

by Eberhard van der Laan,

Mayor of Amsterdam

Research and Innovation:

How to Detect Suspect

Behaviour at Sea

ERCIM NEWSwww.ercim.eu

Number 98 July 2014

SmartCities

Special theme

ERCIM NEWS 98 July 2014

ERCIM News is the magazine of ERCIM. Published quar-

terly, it reports on joint actions of the ERCIM partners, and

aims to reflect the contribution made by ERCIM to the

European Community in Information Technology and

Applied Mathematics. Through short articles and news

items, it provides a forum for the exchange of information

between the institutes and also with the wider scientific com-

munity. This issue has a circulation of about 6,500 printed

copies and is also available online.

ERCIM News is published by ERCIM EEIG

BP 93, F-06902 Sophia Antipolis Cedex, France

Tel: +33 4 9238 5010, E-mail: [email protected]

Director: Jérôme Chailloux

ISSN 0926-4981

Editorial Board:

Central editor:

Peter Kunz, ERCIM office ([email protected])

Local Editors:

Austria: Erwin Schoitsch, ([email protected])

Belgium:Benoît Michel ([email protected])

Cyprus: Ioannis Krikidis ([email protected])

Czech Republic:Michal Haindl ([email protected])

France: Thierry Priol ([email protected])

Germany: Michael Krapp

([email protected])

Greece: Eleni Orphanoudakis ([email protected]),

Artemios Voyiatzis ([email protected])

Hungary: Erzsébet Csuhaj-Varjú ([email protected])

Italy: Carol Peters ([email protected])

Luxembourg: Thomas Tamisier ([email protected])

Norway: Truls Gjestland ([email protected])

Poland: Hung Son Nguyen ([email protected])

Portugal: Joaquim Jorge ([email protected])

Spain: Silvia Abrahão ([email protected])

Sweden: Kersti Hedman ([email protected])

Switzerland: Harry Rudin ([email protected])

The Netherlands: Annette Kik ([email protected])

W3C: Marie-Claire Forgue ([email protected])

Contributions

Contributions should be submitted to the local editor of your

country

Copyright notice

All authors, as identified in each article, retain copyright of

their work

Advertising

For current advertising rates and conditions, see

http://ercim-news.ercim.eu/ or contact [email protected]

ERCIM News online edition

The online edition is published at

http://ercim-news.ercim.eu/

Subscription

Subscribe to ERCIM News by sending an email to

[email protected] or by filling out the form at the

ERCIM News website: http://ercim-news.ercim.eu/

Next issue

October 2014, Special theme: Software Quality

Editorial Information

ERCIM NEWS 98 July 2014 3

Keynote

Smart Cities

Smart Cities are a bit like football: Every city has a teamworking on a “Smart City” and wants to be the Smartest Cityin the world, and at the start of every season every supporterthinks his or her team will be the global champion. Variousranking systems exist, comparing cities on indicators rangingfrom energy consumption per capita to life expectancy; fromWifi coverage to crime-rates. In other words: Smart Citiesare about everything and therefore about nothing. To befrank: I do not really believe all this Smart City marketing. Ido believe that innovation and technology gives us theopportunity to improve the quality of life of the citizens andmake our cities more competitive.

As a mayor of a European capital city, which is ranked rea-sonably high in most European rankings and comparisons, Iwill try to paint a picture of what a Smart City means to meand what it means for our city. As well as the benefits ofSmart Cities, I also see a few risks that sometimes go unrec-ognized. But first a bit of history, because I strongly believethat the DNA of the city will never change and every devel-opment in Amsterdam is based on its (long and prosperous)history.

One could argue that Amsterdam was the first Smart City inthe world; not with the start of our Smart City program in2009, but as early as the end of the 16th century whenAmsterdam began to grow in importance as a trading city.One of the key reasons that Amsterdam became one of therichest cities of that time, with over 50% of all the sea-ves-sels in the world departing and arriving at its harbour, wasthe availability of data on trade and cargo. In a physicalsquare of 400 by 400 meters all the information on the cargo,destination and ownership of all these vessels was gathered.This enabled tradesman to trade cargo that hadn’t even beenoffloaded from the vessel yet and to compare one another’sproducts. The rich data enabled the start of the first stockexchange and the opportunity for everybody to invest intrade, generating investments never seen before. Financialnewspapers shared the information with everybody whowanted it. Essentially, it was a data driven economy withOpen Data avant la lettre.

The sharing of information provided a whole new dimen-sion: “normal” citizens of Amsterdam could take part in thecity’s welfare and economic growth. This is why Amsterdamis not a city of bishops and kings but of citizens and mer-chants, without castles and cathedrals. In Dutch art from the17th century ordinary people were painted; Vermeer'sMilkmaid now displayed in the national museum is a greatexample.

Back to the now: what is the DNA of Amsterdam and howdoes it apply to our approach? In this context I would statethat it is about: openness, entrepreneurship, collaborationand inclusion.

Smart Cities are about openness: open data, open infrastruc-tures and open innovation are at the very heart of the devel-opment of our Smart City. Like the data on cargo and theaccess to the canals, every “Amsterdammer” should have

access to our infrastructure. The sharing of informationdrives innovation because individuals and business can buildon jointly gathered knowledge.

The second subject is entrepreneurship. New technologieswill lead to new business opportunities, new business modelsand new types of companies; companies that are based oncommunities, sharing, or new products and services. Thisincludes companies that have business models that will leadto reduced, rather than increased, energy consumption or carmiles - companies such as Car2Go, Spaces or energy servicecompanies.

The third subject, “Collaboration”, is of course important ona European scale, for example within Eurocities or the SmartCities and European programmes, but it is particularlyimportant that strong collaborations occur between compa-nies, knowledge institutes, government(s) and citizens on alocal level. Everybody is a stakeholder in the developmentof the city and all stakeholders are equally important, whichmeans there are hundreds of thousands of stakeholders.

This brings me to the last point: inclusivity - a city for all“Amsterdammers” and its visitors. All “Amsterdammers”should have the opportunity to be part of the city we areliving in.

I also see many new challenges - challenges relating to pri-vacy, the ability for everybody to benefit from these newtechnologies, and the rapidly changing business which willlead to a large shift in labour. I expect companies, govern-ments and scientists to take responsibility for these chal-lenges and find solutions to address the downside of thesesmart developments.

Having said this I would like to conclude on a simple note: Acity is as great as the people living within it. This will stillhold true in a smart city: it is all about the people who live,work and play there. As a Mayor I will make sure that ourcitizens are always ranked first. This will of course make usthe Johan Cruyff of the world’s Smart Cities.

Eberhard van der Laan

Eberhard van der Laan, Mayor of Amsterdam

Phot

o: S

. van

der

Tor

ren


Contents

2 Editorial Information

KEYNOtE

3 Smart Cities

by Eberhard van der Laan, Mayor ofAmsterdam

JOINt ERCIM ACtIONS

6 ERCIM Fellowship Programme

6 ERCIM 25 Years Celebration

6 W3C and the Web of Things

7 COMPOSE - Converging

the Internet of Services with the

Internet of Things

SPECIAL tHEME

This special theme section “Smart Cities” has been coordinated by IoannisAskoxylakis, ICS-FORTH, Greece, and Theo Tryfonas, Faculty of Engineering,University of Bristol, UK.

Introduction to the Special Theme8 Future Cities and Smart Technologies:

A Landscape of Ambition and Caution

by Theo Tryfonas and Ioannis Askoxylakis

9 Urban Future Outline - A Roadmap on Research for Livable Cities

by Martina Ziefle, Christoph Schneider, Dirk Vallée, Armin Schnettler,Karl-Heinz Krempels and Matthias Jarke

10 Urban Civics - Democratizing Urban Data for Healthy Smart Cities

by Sara Hachem, Valérie Issarny, Alexey Pozdnukhov and Rajiv Bhatia

12 AppCivist - A Service-oriented Software Platform for Social Activism

by James Holston, Rodrigo Ochigame and Animesh Pathak

13 CityLab@Inria - A Lab on Smart Cities fostering

Environmental and Social Sustainability

by Valérie Issarny

14 A Framework for Improving the Multi-Device User Experience in

Smart Cities

by Luca Frosini and Fabio Paternò

16 Moving Towards Interoperable Internet-of-Things Deployments in

Smart Cities

by Gregor Schiele, John Soldatos and Nathalie Mitton

17 Realizing Smart City Scenarios with the ALMANAC and DIMMER

Platforms

by Mark Vinkovits, Marco Jahn and René Reiners

19 Internet of Things Applications for Neighbourhood Embedded Devices

by Joan Fons, Daniel Gaston, Christophe Joubert and Miguel Montesinos

20 Internet of Things: A Challenge for Software Engineering

by Charles Consel and Milan Kabac

22 Semantic Management of Moving Objects in Smart Cities

by Sergio Ilarri, Dragan Stojanovic and Cyril Ray

24 Flexible Access to Services in Smart Cities: Let SHERLOCK Advise

Modern Citizens

by Roberto Yus, Eduardo Mena, Sergio Ilarri and Arantza Illarramendi

25 Quantifying the Benefits of Taxi Trips in New York through

Shareability Networks

by Paolo Santi, Giovanni Resta and Carlo Ratti

26 Integrated Electric Vehicles Sharing and Pooling Mobility Solutions

for Smart Cities

by Marie-Laure Watrinet, Gérald Arnould, Hedi Ayed and DjamelKhadraoui

28 A Carpooling Recommendation System in the Smartphone Age

by Marcell Fehér and Bertalan Forstner


Joint ERCIM Actions

RESEARCH AND INNOVAtION

This section features news aboutresearch activities and innovative developments fromEuropean research institutes

52 Supporting the Design Process of

Networked Control Systems

by Alexander Hanzlik and ErwinKristen

54 How to Detect Suspect Behaviour

at Sea

by Anders Holst

56 A Breakthrough for Balanced

Graph Partitioning

by Fatemeh Rahimian

EVENtS, IN BRIEf

57 Conference and Workshop

Announcements

59 International Innovation Award

for Martin Kersten

59 European Project on Precision

Farming

59 Roberto Scopigno receives the

Eurographics “Distinguished

Career Award” 2014

29 A Smart Parking Campus: An Example of Integrating Different

Parking Sensing Solutions into a Single Scalable System

by Enrique Moguel, Miguel Ángel Preciado and Juan Carlos Preciado

31 Stochastic Travel Planning for Unreliable Public Transportation

Systems

by Tim Nonner, Adi Botea, Marco Laumanns

32 A Quantitative Approach to the Design and Analysis of Collective

Adaptive Systems for Smart Cities

by Maurice ter Beek, Luca Bortolussi, Vincenzo Ciancia, Stefania Gnesi,Jane Hillston, Diego Latella and Mieke Massink

33 Query-Driven Smart Grid City Management

by Stamatis Karnouskos

34 ‘U-Sense’, A Cooperative Sensing System for Monitoring Air Quality

in Urban Areas

by Giuseppe Anastasi, Paolo Bruschi and Francesco Marcelloni

36 Monitoring and Controlling Energy-positive Public Lighting:

The E+grid System

by Balázs Csanád Csáji, Borbála Háy, András Kovács, GianfrancoPedone, Tibor Révész, and József Váncza

37 Demand-Side Management in Smart Micro-Grids: An Optimization

Perspective

by Talbi El-Ghazali

39 Cyber Physical Systems give Life to the Internet of Energy

by Giampaolo Fiorentino and Antonello Corsi

40 When Smart Cities meet Big Data

by Vincenzo Gulisano, Magnus Almgren and Marina Papatriantafilou

41 cItyAM: Managing Big Urban Data for Analyzing and Modelling

Cities

by Alessandro Bozzon, Claudia Hauff and Geert-Jan Houben

43 Urban-Scale Quantitative Visual Analysis

by Josef Sivic and Alexei A. Efros

45 Mobile Augmented Reality Applications for Smart Cities

by Mathieu Razafimahazo, Nabil Layaïda, Pierre Genevès and ThibaudMichel

46 Monitoring People’s Behaviour using Video Analysis and Trajectory

Clustering

by Francois Bremond, Vania Bogorny, Luis Patino, Serhan Cosar, GuidoPusiol and Giuseppe Donatiello

47 Trusted Cells: Ensuring Privacy for the Citizens of Smart Cities

by Nicolas Anciaux, Philippe Bonnet, Luc Bouganim and PhilippePucheral

49 Smart City Operation Center: A Platform to Optimize Urban Service

Rendering

by Filippos Gouidis, Theodore Patkos and Giorgos Flouris

51 Building Smarter Cities through ICT-driven Co-Innovation

by Christophe Ponsard, Robert Viseur and Jean-Christophe Deprez

Joint ERCIM Actions


The 25th ERCIM anniversary and the ERCIM fall meetings will be heldat the CNR Campus in Pisa on 23-24 October 2014.

On the occasion of ERCIM’s 25th anniversary, a special session andpanel discussion will be held on 23 October in the afternoon in the audi-torium of the CNR Campus. Speakers and representatives from research,industry, the European Commission, and the ERCIM community will beinvited to present their views on research and future developments ininformation and communication science and technology.

For more information, please contact:

Adriana Lazzaroni, CNR, ItalyE-mail: [email protected]

ERCIM 25 Years Celebration

ERCIM “Alain Bensoussan”

fellowship Programme

ERCIM offers fellowships for PhD holders from all over the world.

Topics cover most disciplines in Computer Science, InformationTechnology, and Applied Mathematics.

Fellowships are of 12-month duration, spent in one ERCIM memberinstitute.

ConditionsApplicants must:• have obtained a PhD degree during the last 8 years (prior to the appli-

cation deadline) or be in the last year of the thesis work with an out-standing academic record

• be fluent in English• be discharged or get deferment from military service• have completed the PhD before starting the grant (a proof will be

requested).

Application deadlines30 April and 30 September

More information and application form:

http://fellowship.ercim.eu/

W3C and the Web

of things

W3C held a workshop on the Web of Thingsin Berlin on 25-26 June 2014, hosted bySiemens. The Web of Things is expected tohave broad and sweeping economic andsocietal impact. Open standards will be crit-ical to enabling exponential growth of thekind we experienced with the early days ofthe Web.

Until now, the focus has been on the devicesand communications protocols. There is agrowing awareness that the business oppor-tunities will be centered on the associatedservices, moreover, the current situation isone of fragmentation with products beingdeveloped in isolation due to a plethora ofIoT protocols and a lack of a sharedapproach to services.

The workshop examined the opportunitiesfor open Web standards for service plat-forms in the network edge (e.g. home gate-ways) and the cloud, along with the chal-lenges for security, privacy and the integra-tion with the Web of data. A workshopreport is now in preparation. This workshopwas supported by the EU project Compose .

Links:

Compose project: http://www.compose-project.eu/WoT workshop:http://www.w3.org/2014/02/wot/


COMPOSE - Converging

the Internet of Services

with the Internet of things

COMPOSE (Collaborative Open Market to Place Objects

at your Service) will create an ecosystem for unleashing

the power of the Internet of Things (IoT) via an easy

transformation to an Internet of Services (IoS). The

technology developed by the project will enable the

creation of base services, composite services, and

applications stemming from and operating on smart

objects.

COMPOSE aims at enabling new services that can seam-lessly integrate real and virtual worlds through the conver-gence of the Internet of Services (IoS) with the Internet ofThings. COMPOSE will achieve this through the provi-sioning of an open and scalable marketplace infrastructure,in which smart objects are associated to services that can becombined, managed, and integrated in a standardised way toeasily and quickly build innovative applications. The projectwill develop novel approaches for virtualising smart objectsinto services and for managing their interactions. This willinclude solutions for managing knowledge derivation, secureand privacy- preserving data aggregation and distribution,dynamic service composition, advertising, discovering, pro-visioning, and monitoring.

COMPOSE is expected to give birth to a new businessecosystem, building on the convergence of the IoS with theIoT and the Internet of Content (IoC). The COMPOSE mar-ketplace will allow SMEs and innovators to introduce newIoT-enabled services and applications to the market in a shorttime and with limited upfront investment. At the same time,major ICT players, particularly cloud service providers andtelecommunications companies, will be able to repositionthemselves within a new IoT- enabled value chain.

Technical ApproachThe vision of the COMPOSE project is to advance the stateof the art by integrating the IoT and the IoC with the IoSthrough an open marketplace, in which data from Internet-connected objects can be easily published, shared, and inte-grated into services and applications. The marketplace willprovide all the necessary technological enablers coveringboth delivery and management aspects of objects, services,and their integration:• Object virtualization to enabling the creation of standard-

ized service objects• Interaction virtualization - abstract heterogeneity while

offering several interaction paradigms• Knowledge aggregation to create information from data• Discovery and advertisement of semantically-enriched

objects and services• Data Management to handle massive amounts and diversi-

ty of data/metadata• Ad hoc creation, composition, and maintenance of service

objects and services• Security, heterogeneity, scalability, and resiliency, incor-

porated throughout the layers

Expected ImpactCOMPOSE strives for a strong impact on a developingmarket by lowering barriers to develop, select, combine, anduse IoT-based standardized value added services. This willbe achieved by providing a complete ecosystem, and havingit adopted by enterprises, SMEs, government-related bodies,and individual developers and end-users.

We hope that opening the door to this realm for smaller enti-ties will lead to higher innovation. COMPOSE expects to aidby fostering a developers' community and advocating anopen source/interfaces

Use-CasesCOMPOSE design, development, and validation will bebased on innovative use cases highlighting different aspectsof the platform: • Smart Shopping Spaces: this use cases will pilot COM-

POSE in shopping environments, focusing on the dynam-ic composition and delivery of services starting from prod-ucts available in shops.

• Smart City (Barcelona): Ample amount and diversity ofsensors are deployed at a Barcelona district under thesupervision of a COMPOSE partner. Along withBarcelona's OpenData, COMPOSE intends to showcaselife in a smart city by creating a group of city services forthe citizens.

• Smart Territory (Trentino): With the collaboration ofregional network providers, the tourism board, and mete-orological data providers, COMPOSE will explore inno-vative services for tourists. This pilot aims to enhance thetourist experience by exploiting COMPOSE technologiesfor the creation of personalized, social- and environmen-tally-aware (web and mobile) tourism services and territo-ry monitoring services that leverage the regional network-ing and environmental infrastructures

Survey on promising IoT Application DomainsThe project is currently collecting useful insights aboutexisting and future exploitation opportunities in the IoT fieldfor the expected project outcomes. The questionnaire is tar-geted at industry and technology experts and asks to identify,on the basis of their experience, the most promising IoTapplication domains and then specify them in terms of: (i)existing market drivers and inhibitors and (ii) success storiesand associated business models. Eventually, with respect tothe outlined context, experts are kindly asked to provide theirinsights about the most promising features of the COMPOSEproject. The questionnaire is available on the project website.

W3C is a partner of the project through ERCIM EEIG. A taskof W3C is to facilitate standardization and exploitation of theactivities. Fraunhofer FOKUS the project comprises 12 part-ners from industry and academia as well as standardizationbodies, including the ERCIM member Fraunhofer FOKUS.

Link: http://www.compose-project.eu

Please contact:

Philippe Hoschka, W3CE-mail [email protected]

7


Special theme: Smart Cities

Introduction to the Special Theme

future Cities and Smart technologies:

A Landscape of Ambition and Caution

by Theo Tryfonas and Ioannis Askoxylakis

The old Chinese curse of “may you live in inter-esting times” has never been more relevant thanin the field of urban development. For some timenow, the global urban population has exceededthe global rural population. Cities and cityregions have therefore emerged as the onlygrowth models capable of meeting the increaseddemands and strains facing global supply sys-tems, which are seriously affected by populationgrowth, climate change, globalization and inter-national security issues. Given these constraints,urban development is becoming a tough chal-lenge.

Smart technologies may play a role in addressingthese issues. From sensor networks to wirelessconnectivity and big data analytics to the Internetof Things, our planners, architects, infrastructureengineers and facilities managers have theability to harness the power of information gen-erated within the boundaries of cities and thebuilt environment. Transport planners can betterunderstand the demand for services and pas-senger behaviour; similarly, energy providerscan tailor their supply to the real needs of theircustomers. End users can benefit from personal-ized and timely service provision that takes intoaccount the individual’s location and situation -from way finding to micropayments to finding agood place to dine. There is enormous potentialto use these technologies to establish more sus-tainable behaviour: for example, with energy usemonitoring informed by smart meters, personal-ized transport assisted by e-ticketing and mobiletechnology, wireless infrastructure, to name justa few possibilities.

This capability, however, comes at a significantcost. Real time monitoring and bulk data collec-tion call for high performance computationalinfrastructure, huge storage capacity, widereaching connectivity, digital skills and a posi-tive governance attitude. Furthermore, this tech-nology creates the potential for mass surveil-lance operations by the state as well as globalcorporations, triggering grave concerns for indi-viduals’ privacy. This is a promising yet chal-lenging field that has a significant role to play increating a more sustainable urban future.Technologies deployed and tested across mul-tiple cities provide insights into a future that we

may wish to see, but also give us forewarningsabout the challenges that lie ahead.

This special issue of ERCIM News captures thestate of art of European research and provides aninsight into smart systems and technologies thatare already, or may become in the future, part ofour daily urban lives. Although it is difficult toprovide an exact definition for it so early in itsdevelopment, these technological advances con-tribute fundamentally to a forward-thinkingconcept that has been termed ‘Smart City’, i.e.,an urban environment that capitalizes on con-temporary information management capabili-ties, including ICT infrastructure and applica-tions, to improve on the delivery of public serv-ices, transport, health, sustainability, theeconomy and the overall wellbeing of its resi-dents. A recent report from the Department ofBusiness, Innovation and Skills (BIS) in the UKestimated the emerging global market of SmartCities to worth a minimum of $400 billion. Asan emerging domain, however, there are still anumber of uncertainties to be addressed.

With a wide selection of articles from novel net-working technologies to digital entrepreneurialprocesses, it is clear from this special issue thatno single technology or vendor can deliver thepromises of the Smart City. No amount of sen-sors or data will offer a solution without pur-poseful assembly of the right components,including technology, skills and governance.This is an important perspective which high-lights the need for European policy, systemarchitecture and systems integration to beinformed by the challenges of understandinghuman behaviour, fostering creativity by devel-oping the human capital and ensuring compli-ance with the data protection frameworkthrough balancing the need for security withrespect for individual rights. What interestingtimes, indeed!

Please contact:

Ioannis AskoxylakisICS-FORTH, GreeceE-mail: [email protected]

Theo Tryfonas, Faculty of Engineering, University of BristolE-mail: [email protected]


One of the key challenges for contempo-rary research is to develop and maintainlivable, sustainable and resilient cities.For the first time in history, the majorityof the world’s population live in urbanspaces. The complexity of urban areasnecessitates interdisciplinary approachesto be taken in research and development.Goods and services have to be availableat short notice. Mobility must be afford-able and reliable. Technical infrastruc-ture needs to be easily accessible.Citizens require appealing living quar-ters embedded in open spaces with greenand blue infrastructure. Often, thesedemands are contradictory, thusimposing considerable challenges onurban planners.

UFO is organized in three subprojects:Future Mobility refers to sustainableinfrastructure; Future Energy tacklesquestions regarding energy supply andthe transition to renewable energy pro-duction; and the fields of biodiversityconservation and climate change adapta-tion are the targets of Future Ecology.

Future Mobility Mobility is an essential prerequisite forpeople participating in social and eco-nomic lifes. In light of the peak oildebate, the EU guidelines for air pollu-tion control and environmental noiseand demographic developments,mobility concepts must change drasti-cally. Transport systems need to behighly adaptive to periods of peak usage(i.e., rush hours) at multiple scales (e.g.,day-by-day or seasonally). They mustmeet a range of community needsincluding accessibility, comfort, safety,sustainability, affordability and environ-mental justice. Mobility options must beintermodal, flexible and designed as“door-to-door” mobility chains. Thepredominantly technology-centeredplanning of infrastructural mobility con-cepts, without integrating citizens into

the decision making processes, is nolonger viable.

This sub-project combines city andmobility planning with the technicalpossibilities of information and com-munication, the cognitive and commu-nicative requirements of urban popula-tions, discursive practices, media strate-gies and gender aspects. The goal is todevelop a multifactorial informationand communication strategy thatinvolves the social, individual, commu-nicative and cognitive needs of the citi-zens, as well as technical and urbanplanning related aspects of feasibility.

Future Energy This sub-project focuses on energyturnaround and the interrelationbetween the technical, economical, IT,ecological and social perspectives. Itsaim is to develop a holistic model and amethodology for the implementation ofsustainable, robust energy systems thatsystematically include social factors(i.e., user perceptions of energy sys-tems) into the identification, planning

and realization of energy scenarios.Within the context of environmentaland technical conditions, acceptance-relevant factors (perceived as the bene-fits and drawbacks) from differentregional contexts are collected, evalu-ated and modeled to determine theirrelationships. The inclusion of socialknowledge is accomplished via threedata sources: the empirical modelling ofcognitive-affective attitudes, theanalysis of opinion-forming processeson the Internet and an assessment ofenvironment-related aspects. Potentialdecision trade-offs are identified usingconjoint analyses.

The results are integrated into the devel-opment of technical, economical and ITtransformation processes. Based on thismodelling, an understanding on thetechnical, economical and social per-spectives on energy transition arederived. These results are useful forpoliticians and decision makers andmake an important contribution indeveloping transparent and adequatecommunication policies.

Urban future Outline -

A Roadmap on Research for Livable Cities

by Martina Ziefle, Christoph Schneider, Dirk Vallée, Armin Schnettler, Karl-Heinz Krempels and Matthias Jarke

Urban Future Outline (UFO), is a project funded by the Excellence Initiative of the German states and

federal governments at RWTH Aachen University (2013-2015). It aims at developing a holistic

approach, in which energy needs, mobility demands, ecological and climatological requirements and

human demands are addressed, against the background of socially responsible technology

development in urban areas.

Figure 1: The overall approach and the topics addressed within UFO

Future Ecosystem Planning, implementing and sustaininglivable urban spaces that pay equal atten-tion to humans, technology and nature isa task that has not been solved holisti-cally to date. Situations where combinedstressors are acting pose specific chal-lenges for urban development becausethe negative effects of superimposedstresses such as heat, noise and air pollu-tion cannot be easily discerned whentaking a single stressor perspective. Thisdeficiency is all the more sensitive at cer-tain times, for example summer, whenthermic differences are more extremeand higher levels of air pollution areexpected due to climate change.

This sub-project aims to (1) considercombined stress situations in a measure-ment and model chain, (2) derive com-bined stress indices and (3) make theseaccessible in a virtual environment(aixCAVE). Apart from the thermiccomponent, air quality, acoustic percep-tion and user perception are all consid-ered. Which mix of combined stressesfor urban residents inhabiting differentlydesigned open spaces will be investi-gated, as will future scenarios. Asopposed to previous approaches to mod-elling the effects of thermic and actinicstresses, (4) stresses will be diversifiedby user profiles, gender, and age.

Overall, the research design in UFOaims to make use of the most recentdevelopments in technology, informa-tion sciences and natural sciences. Atthe same time, it sought to develop afeasible solution to integrating thesocial, economic and cultural needsassociated with urban development [1].An integral part of the UFO project,therefore, is the multi-disciplinaryapproach [2] [3]. The consortium con-sists of nine scholars from RWTHAachen University:• Christoph Schneider - Physical

Geography • Martina Ziefle - Communication Sci-

ence • Dirk Vallée - Urban and Transport

Planning• Matthias Jarke - Information Systems• Armin Schnettler - High Voltage

Technology• Eva-Maria Jakobs - Textlinguistics • Carmen Leicht-Scholten - Gender

and Diversity • Janina Fels - Technical Acoustics• Andreas Schäffer - Environmental

Biology • Peter Russell - Computer Aided

Architectural Design• Thomas Niehr - German Linguistics• Thorsten Kuhlen - Virtual Reality

References:

[1] C. Schneider, B. Achilles, H.Merbitz: “Urbanity and Urbanization:An Interdisciplinary ReviewCombining Cultural and PhysicalApproaches”, Land 2014, 3(1), 105-130

[2] S. Krengel, T. Falke, A. Schnettler:“Optimization Model for the EnergySupply in City Quarters”, in proc. ofCIRED 2013, Stockholm

[3] S. Himmel, M Ziefle, K. Arning,:“From Living Space to Urban Quarter:Acceptance of ICT MonitoringSolutions in an Ageing Society”, in M.Kuroso (ed.), Human-ComputerInteraction, Users and Contexts of Use.Springer LNCS 8006, pp. 49-58.

Please contact:

Matthias Jarke, Fraunhofer-FIT, GermanyE-mail:[email protected]

Martina Ziefle, Christoph SchneiderRWTH Aachen, GermanyE-mail: [email protected],[email protected]



Urban Civics - Democratizing Urban Data

for Healthy Smart Cities

by Sara Hachem, Valérie Issarny, Alexey Pozdnukhov and Rajiv Bhatia

The Urban Civics project brings together a multi-disciplinary, trans-atlantic team of experts working

towards a common goal that citizens and governments collaborate in achieving participatory democracy

in healthy cities.

Technologies that bring together theemerging visions for smart, connectedand resource-efficient cities, and the rev-olutions in social media and mobile net-working, offer unprecedented opportu-nities for participatory democracy,health and sustainability. Early propo-nents of Smart Cities have advanced theidea of hyper-connected city informa-tion systems as a means of promotingoperational efficiency and naturalresources conservation. However, theidea of citizen participation in the designand control of these systems has beenmissing. In democratic political sys-

tems, optimization in cities also meansbeing responsive to the observations,attitudes and demands of their citizens.

The widespread adoption of social net-working and mobile communication, aswell as the Internet-of-Things (IoT), cit-izen-science and open data, generatenew information resources for under-standing the needs and preferences ofcitizens from government services.Information generated via citizen sci-ence and crowd sensing sources, are par-ticularly good at reflecting the prioritiesand perceptions of citizens. Inclusive

and participatory, “smarter” cities canintegrate and leverage these diverseinformation resources to improveresponsiveness to citizens and improvethe urban quality of life. For example,environmental conflicts such as air andnoise pollution can be better managed.Still, supporting participatory informa-tion gathering at urban scales, and inte-grating that sensed data into informationthat is relevant to urban governance andmoreover, actionable, remains a majorresearch issue. The Urban Civics projecttackles this challenge by bringingtogether a multi-disciplinary team of


experts, so that citizens and govern-ments may work together towardhealthy cities.

In a nutshell, Urban Civics is devel-oping, from the design to prototypeimplementation, a middleware solutionthat tackles the following research ques-tion (Figure 1): ‘How can urban datasources be leveraged and comprehen-sively integrated to accurately capturenuisance issues facing our cities?’. Wehave specifically concentrated onaddressing the following associatedresearch challenges:

• How can we develop a distributedsystem architecture that leverages therichness of the urban sensors of thenew digital era (featuring IoT, opendata, citizen science, social network-ing and mobile computing) at large

scales? Our approach lies in leverag-ing relevant probabilistic protocols toovercome the urban scale challengestogether with building upon semanticWeb technologies to aggregate physi-cal and social data sources [1].

• How can we ensure citizen participa-tion? Our approach lies in studyingdedicated machine-learning algo-rithms (e.g., see [2]), to mine leader-ship and thereby prompt citizen par-ticipation, as well as optimizing datacollection strategies through incen-tive-driven, pro-active citizen engage-ment and a more informed approachto crowd-sensing.

• How can we assimilate the rich urbandata to develop significant city mod-

els? We address this challenge bybuilding on the assimilation of obser-vations that deal with time-varyingdistributions, and take into accountboth model and observational errors toproduce the most accurate map of themonitored phenomenon (e.g., see [3]).

The Urban Civics middleware solutionapplies to a variety of environmentalissues that our cities are currently facing.Indeed, the physical and social sensingapproach implemented by Urban Civicsremains similar across use-cases dedi-cated to environment monitoring.Specifically, the approach requires plug-ging dedicated sensors to provide quan-titative data and applications to enableend-users to provide qualitative data,both of which are aggregated throughdata modelling & assimilation function-alities. The mathematical details vary

depending on the use-case specific phe-nomenon. As an illustration, we discussnoise pollution and air quality below:

• Noise pollution, i.e., the undesiredsounds that harm one’s well being,come from several sources includingtraffic, neighbors and constructionworks. Noise can be measured using amicrophone, hosted on sound/noiselevel meters which can be deployedthroughout a city. However, whilsthighly accurate, static sensors alonecannot account for the fine-grainedspatial variations that occur.. Conse-quently, smartphone microphones canbe exploited as a complementary datasource; end-users describe noise lev-els through a dedicated applicationplugged in with Urban Civics. Addi-

tional information can also be auto-matically extracted from social net-works, e.g., if users are attending apublic event at the measurement time.Once the various inputs are provided,the data is analyzed and proper citymodels (of noise pollution) are gener-ated.

• Air quality monitoring, in particularthe tracking of nitrogen dioxide(which is mostly emitted from vehi-cles), can be performed by deployingstatic sensors throughout a city. Fur-ther, since smartphones do not yethost sensors that can monitor air pol-lution, end-users can themselves beequipped with such sensors and canprovide their own qualitative input todescribe air quality at their locations.Data assimilation for air quality isalready applied in Paris and leveragedin several cities in France. However,urban-scale air quality monitoring isstill challenged by the requireddeployment of sensors, which is mademuch more tractable using UrbanCivics and its diverse urban sensors.

Urban Civics is under active develop-ment, as part of larger research initia-tives oriented toward Smart Cities thatare being launched by both University ofCalifornia, Berkeley, and Inria. The ini-tiatives will promote collaborationbetween the two organisations. In par-allel, we are working closely with thecity administration to acquire insights onthe broad applicability of Urban Civics.

Link:

CityLab@Inria Project Lab on SmartCities: http://citylab.Inria.fr

References:

[1] S. Hachem, A. Pathak, V. Issarny:“Service-oriented middleware for large-scale mobile participatory sensing”,Pervasive and Mobile Computing, 2014 [2] F. Kling, A. Pozdnoukhov: “When acity tells a story: urban topic analysis”,in proc. of the 20th InternationalConference on Advances in GeographicInformation Systems, 2012[3] A. Tilloy et al.: “BLUE-based NO2data assimilation at urban scale”,Journal of Geophysical Research,118(4), 2013.

Please contact:

Valerie Issarny, Inria@Silicon Valley, France, USAE-mail: [email protected]

Figure 1: The role of Urban Civics in urban democracy

The increased adoption of mobiledevices and social networking mobilizescitizens to monitor their living environ-ment actively. Ultimately this engage-ment is likely to prompt governments totake action. Therefore, the study ofsocially-grounded software systems thatare dedicated to making cities bothsmarter and more inclusive holds greatpotential for cross-domain researchamong social scientists and computerscientists.

In this context, mobile social net-working appears as a significant tool tosupport social activism within cities.However, the development of informa-

tion and communication technologies(ICT) for social activism is still in itsinfancy, ranging from the use of emailfor internal communication to the Weband social media for mass communica-tion. In terms of integrated platforms,there have been a few initial offeringssuch as Google Apps for Non Profits.However, Google Apps is a generic plat-form for organizations and requires sig-nificant customization to meet the needsof social activists. At the same time, ini-tiatives like the Social Apps Lab atCITRIS at the University of California,Berkeley, clearly show the potential ofmobile devices to generate an active,critical and direct engagement of citi-

zens in social issues, which mayincrease the impact of social activism.

The AppCivist VisionAppCivist is a software system beingdeveloped by the Social Apps Lab inpartnership with Inria, France. It pro-vides a platform for democraticassembly and collective action that letsusers make their own applications,called Assemblies, with modular com-ponents. An underlying premise of theproject is that the development of ICTsfor social activism crucially depends onthe provision of adequate software toolsso that activists can easily assembleapplications matching their needs. From

a software engineering perspective,architectural principles like service-ori-entation, coupled with the latestadvances in distributed computing,appear as relevant base building blocks.From the social science perspective,social activism relies on a number ofkey functionalities that may be lever-aged and re-used effectively acrossapplications. Such functionalitiesinclude the following, among others:• Proposing, deliberating, and voting

allow citizens to have meaningfuldiscussions based on available infor-mation, leading to action. While sim-ple plurality-based voting systemsexist, they fail to enable the nuanced

discussion, filtering, consideration,and consensus-building that is funda-mental to both social activism anddemocratic citizenship.

• Collaborative mapping helps peoplecrowd-source the creation of custommaps that display phenomena of theirinterest. The Social Apps Lab hasalready successfully deployed thisfunctionality as part of the DengueTorpedo project in the favelas of Riode Janeiro, Brazil.

• Data visualization enables non-spe-cialist users to gather and convertinformation quickly into easily-con-sumable graphics and reports forwide dissemination.

• Mobilizing citizens encourages peo-ple to specify what needs to be doneand to get involved. AppCivist helpsusers raise issues, propose solutions,and evaluate proposals. As activismdepends on follow-through, it alsotracks the progress of activities,allows citizens to collaborate, andrewards those who participate.

• Linkage with media and administra-tion enables activists to report issuesto those in power and to track theirresolution.

These functionalities encourage pro-posal-making that is more directly dem-ocratic and enable people to make betterarguments for their proposals. Makingbetter arguments is key to advancingdemocratic deliberation. At the sametime, the field of social activism bringsunique challenges to service-orientedcomputing in particular and to ICT ingeneral, including the following:• Focus on Usability. Since the users of

AppCivist are not expected to be ICTexperts or “digital natives”, it isimperative that the use of the plat-form’s modular functionalities beintuitive. The same applies to theadministration of plug-and-play andinterdependent individual services.

• Cost-sensitivity. The platform shouldaccommodate the vastly differentbudgets of activist organizations.Consequently, it should carefully



AppCivist - A Service-oriented Software

Platform for Social Activism

by James Holston, Rodrigo Ochigame, and Animesh Pathak

Anthropologists and ICT researchers from France and USA join forces to enable

social activists to leverage the tools of the digital revolution.

Figure 1: AppCivist core platform and associated services for social activism.

consider the use of costly servicessuch as SMS gateways and incorpo-rate ways to define and manage budg-ets.

• Privacy. Since activist organizationsand their members may have signifi-cant security issues, AppCivist mustprovide strong guarantees regardingthe privacy of the data and identity ofusers.

• Accountability. Nevertheless, certainactivities (such as proposing anddeliberating) may require that peoplebe accountable for their inputs.Therefore, AppCivist should alsoprovide strong guarantees for non-repudiation, where needed.

• Inclusiveness. We do not want the useof AppCivist to lead to exclusion -something that can happen, for exam-ple, if the platform restricts access tomessaging to those with smart-phones. Consequently, the platformshould enable heterogeneity in differ-ent dimensions of communication,such as language, technology, and soforth.

Next Steps: Enabling RapidIntegration of ICT in Social ActivismIn view of the above, the objective of ourresearch is to study, from design to proto-type implementation, a service-orientedarchitecture and supporting platformaimed at facilitating the assembly ofapplications by social activists. Jointlyperformed at Inria and the Social AppsLab at the University of California,Berkeley, this research will also involveclose consultation with activists.

As illustrated in Figure 1, the develop-ment of the core platform and associatedservices for social activism will includealgorithms and protocols for filteringand voting, data visualization, assemblyof modular components, real-timemulti-platform communication betweenactivists and among the general public,mobile-based crowd-sourcing/sensing,and dynamic task-allocation and man-agement among participants. Further,the interplay between mobile and cloudcomputing entails assessing both costand social effectiveness.

Links:

Social Apps Lab at CITRIS:http://socialappslab.com/ CityLab@Inria Project Lab on SmartCities: http://citylab.inria.fr/

References:

[1] J Holston: “Insurgent Citizenship:Disjunctions of Democracy andModernity in Brazil”, PrincetonUniversity Press, 2008.[2] A. Toninelli, A. Pathak, V. Issarny:“Yarta: A Middleware for ManagingMobile Social Ecosystems”, GPC2011.

Please contact:

James HolstonUniversity of California, Berkeley, USA E-mail: [email protected]

Animesh PathakInria, FranceE-mail: [email protected]


Our Motivation The world is in the midst of animmense population shift, as peoplemove from rural to urban areas. Thishas led governments, businesses andcommunities to rely on technologies, inparticular, information and communi-cation technologies (ICT), to overcomethe challenges posed by this rapidurbanization. As a result, various aca-demic, industry and city-led ICT initia-tives have been launched in recentyears with a view to building “smarturban infrastructures”. These providedetailed information about the func-tioning of a city to its citizens and busi-nesses, thereby enabling them to betterunderstand their infrastructures and

resources and thus, improve the man-agement of them.

While environmental and economicalsustainability have been on the ICTresearch agenda for some time, theequally important social sustainabilityhas been overlooked in the context ofSmart Cities. Indeed, cities are first andforemost places for people, and thusbuilding cohesive, inclusive and flour-ishing communities should be at theforefront of our research agenda.Without the right social infrastructure inplace, problems such as isolation,mental health problems, anti-socialbehaviours and crime are more likely toarise, pushing communities into decline.

Research Themes and Challenges The objective of the CityLab@Inria is tostudy ICT-based Smart City systemsfrom supporting “sensing” systems up toadvanced data analytics and new serv-ices for the citizens that promote socialand environmental sustainability.

Specifically, CityLab@Inria bringstogether Inria project teams in net-working (FUN and URBANET), distrib-uted software systems (ARLES-MiMoveand MYRIADS), data management(DICE, OAK and SMIS) and data ana-lytics (CLIME and WILLOW) to inves-tigate the following research questions: • How can urban-scale sensing that

needs to combine both physical and

CityLab@Inria - A Lab on Smart Cities

fostering Environmental and Social Sustainability

by Valérie Issarny

The Inria Project Lab CityLab@Inria which is currently under creation, studies information and

communications technology (ICT) solutions that promote social and environmental sustainability and

facilitate the transition to Smart Cities. The Lab places a strong emphasis on multi-disciplinary

research, integrating relevant scientific and technology studies from sensing up to analytics and

advanced applications. The idea is that the research environment will mirror the predicted Smart City

Systems of Systems. A central concern of the Lab is running experiments so that we are able to

investigate proposed approaches in real-life settings.

social sensing be effectively sustainedwhile accounting for the requirementsassociated with the target network?These include scalability, energy-effi-ciency and privacy preservation.Sensing the ‘city pulse’ brings chal-lenges for the supporting data man-agement which must scale-up, as wellas integrate highly heterogeneous dataof various qualities. The literature isrich with papers addressing these con-cerns individually. However, they areseldom tackled together, especiallywhile simultaneously considering theurban scale. Our approach to over-come these challenges lies in thestudy of scalable protocols from thenetworking up to the middleware lay-ers, together with advanced tech-niques for privacy enhancement andsemantic-aware data management.

• How can the data be aggregated so thatthe evolution of a city can be not onlyunderstood, but also anticipated andperhaps even influenced? Data analyt-ics is at the core of Smart Cities, mak-ing big data available to us throughsensing. Based on the open data trend,this can become very useful in provid-ing knowledge on the cities. It is a veryactive area of research. However,numerous open problems remainregarding how large-scale data is ana-lyzed and the uncertainty associatedwith urban-scale, crowd-sourced data

collection must also be overcome. Ourcontribution in this area leveragesadvanced research results on dataassimilation and machine learning.

• While city-scale sensing and dataanalytics are two complementaryaspects of Smart City systems, theyare also inter-related as one may ade-quately inform the design of theother. Therefore, it is essential todesign crosscutting architectures forSmart City systems based on thecomprehensive integration of the cus-tom data sensing and analytics thatwe will investigate.

• Finally, the Smart City vision willonly come true if it is accompaniedby concrete urban services that domake our (future) cities sustainableand agile. A number of applicationareas have been suggested and theseinclude smart energy, smart healthand smart transportation. However,we are still lacking disruptive servic-es that will contribute to making ourcities better places to live while alsoaddressing the central challenge ofgrowth. One important question ishow the use of city-scale sensing canimpact city governance, particularlyits social dimension? Our researchwill be guided by the study of newurban services which will be under-taken in close collaboration withexternal partners (especially city rep-

resentatives) as well as researchersfrom the social science field.

While the scientific focus ofCityLab@Inria is broad, the Lab’sresearch leverages relevant effort withinInria project-teams that is further revis-ited as well as integrated to meet thechallenges of smart cities. An International Lab

CityLab@Inria research builds on thecollaborative effort of the internationalresearch community, especially theInria@SiliconValley program. Indeed, akey characteristic of the CityLab@InriaLab is its international dimension whichbegan with the Paris-San Franciscocooperation agreement toward smartercities . This agreement, signed onMarch 20, 2013, is dedicated to devel-oping smarter cities and includes sup-port for targeted research programsamong which is the Joint Inria-CITRISCityLabs Program.

Links:

CityLab@Inria Project Lab on SmartCities: http://citylab.Inria.fr Inria@Silicon Valley program:https://project.inria.fr/siliconvalley/

Please contact:

Valérie Issarny, Inria@Silicon Valley,E-mail: [email protected]



A framework for Improving the Multi-Device

User Experience in Smart Cities

by Luca Frosini and Fabio Paternò

The current evolution of pervasive technologies in our cities means that traditional access

modalities are not always suitable for a particular service. Consequently novel solutions are needed.

Freely moving citizens in urban environments would often like to be able to better exploit available

devices to access services. In such cases, a multi-device user interface in which interactive

components can be dynamically distributed over devices would be extremely useful: unfortunately,

this is not currently supported by available development toolkits. Here, we propose a solution that

extends existing Service Front Ends for Smart Cities and supports multi-device interaction by

exploiting personal devices and public displays.

Our life is becoming a multi-device userexperience. In the last decade, a widevariety of interactive devices have pene-trated the mass market and people arespending increasing amounts of timeusing them. There are two modes ofusage, sequential or simultaneous. Thesequential user moves from one deviceto another at different times to accom-

plish a task whilst the simultaneous userinteracts with more than one device atthe same time, for either related or unre-lated activities [1]. Recent technolog-ical advances have meant that large dis-play screens are now available on themass market and these greatly enhanceusage. Given their low cost, suchdevices are being installed in great

numbers in many public places (e.g.,train stations, airports, hospitals, publicoffices, museums, universities, shopcenters, bars and restaurants). Thus,deployment is occurring in both outdoorand indoor environments, providingvarious types of content (e.g., informa-tive, entertainment or advertisements)[2]. Unfortunately, insufficient attention

application is a set of games for mul-tiple users who can participate with thesupport of any device, either mobile orstationary.

The framework we propose can be usedto easily obtain interactive multi-deviceapplications for different domains andin different contexts in Smart Citiesincluding single or multi-user applica-tions, indoor or outdoor environmentsand mobile and stationary devices.

Links:

HIIS Lab: http://giove.isti.cnr.it Framework available at:http://giove.isti.cnr.it/tools/MUDUIDME/homeIUSDM Project:http://giove.isti.cnr.it/IUDSM

References:

[1] Google: “The new multi-screenworld: Understanding cross-platformconsumer behavior”, Technical report,August 2012,http://www.google.com/think/research-studies/the-new-multi-screen-world-study.html[2] J. Müller, F. Alt, A. Schmidt, D.Michelis: “Requirements and DesignSpace for Interactive Public Displays”,ACM Multimedia, 1285–1294 (2010)[3] S. Hosio et al.: “Supportingdistributed private and public userinterfaces in urban environments”, inproc. of HotMobile 2010, 25-30.

Please contact:

Luca Frosini, Fabio PaternòCNR-ISTI, Italy.E-mail: {luca.frosini,fabio.paterno}@isti.cnr.it


has been paid to how information is pro-vided accessed and through suchdevices, thus diminishing their potentialeffectiveness.

People would like to be able to betterexploit available technology whenusing multiple devices to interact withtheir applications. For example, usersmay wish to dynamically move compo-nents of their interactive applicationsacross different devices with variousinteraction resources [3]. However,managing information across devices ischallenging. Unfortunately, the devel-opment of multi-device user interfaces(UIs) is limited by the interaction devel-opment toolkits currently available.These toolkits are still designed underthe assumption that they are supportingthe development of UIs for singledevices, rather than providing supportfor multi-device access. Consequently,the main issues with multi-device UIsare their poor adaptation to their usagecontext, the lack of coordination amongtasks performed through differentdevices and inadequate support forseamless cross-device task perform-ance.

To address this gap we have designedand developed a novel frameworkwhich is capable of:• providing developers with an API

that can be exploited by both Weband Android applications in order toobtain application UIs that can bemore easily distributed dynamicallyand/or migrated in multi-device andmulti-user environments,

• dynamically creating sessionsbetween groups of users/devices witha distributed UI. The elements of the

Figure 1: An example of a supported multi-device user interface, where a) shows the application providing content in the tourist version and

b)shows how the application can be used to exploit public displays.

UI can be distributed according tospecific device(s), group(s) ofdevices, specific user(s) or groups ofusers classified by roles, and

• avoiding the need for a fixed server tomanage the distribution, which can beuseful when connectivity is limited.

This framework is publicly available athttp://giove.isti.cnr.it/tools/MUDUIDME/home. It provides developers with asmall set of commands that allow themto easily support dynamic situations inwhich flexible UI distribution and goodperformance are requested, with limitedimpact on the code.

Various applications have already beendeveloped using this framework. Oneapplication was designed to improve theuser experience during a museum visit.A single user application which was ableto exploit mobile devices, in conjunctionwith a public display in an indoor envi-ronment, was implemented. Themuseum has some large public displays,which allow visitors to access multi-media content. When users are near thelarge screens, they can use their smart-phone to select content of interest. Forexample, visitors can select and displaysome high-resolution images of art-works that cannot be viewed from closeup (e.g., because of security or artpreservation issues). Another applica-tion supports city guides who areaccompanying groups, with eithertablets or smartphones. The applicationshows information supporting themobile visit and allows the visitor toselect what content is to be shown in theversion of the tourist role (Figure 1a) orexploit the public displays theyencounter (Figure 1 b). Yet another

http://giove.isti.cnr.it/tools/MUDUIDME/home

Using the Internet-of-Things (IoT) todeploy Internet-Connected-Objects(ICOs) in Smart Cities raise severalchallenges, especially in terms of inter-operability. Drawing on semantic inter-operability solutions developed for IoTsystems, the VITAL project aimed tobridge the numerous silos which exist inIoT deployments in Smart Cities. Thiswas achieved by repurposing andreusing sensors and data streams acrossmultiple applications, without compro-mising citizens’ security and privacy.This approach promises of increase

Return-On-Investment (ROI) figureswhich are associated with the typicallycostly development of Smart City infra-structures by expanding the number andscope of potential applications.

IoT technology enables a large numberof physical and virtual ICOs to be coor-dinated so as to provide human-centricservices in a variety of sectors, especiallyin Smart Cities [1]. Many research, pilotand commercial applications have beendeveloped thus far, ranging from thosewhich use a RFID/Wireless Sensor

Network to those which involve largenumbers of different devices and ICOs.Collectively, these applications have asignificant impact on both business andsociety. Thus, the IoT paradigm is rele-vant to the development and implemen-tation of sustainable urban developmentpolicies. However, a number of IoT sys-tems have been created in parallel whichis leading to the creation of IoT applica-tion silos within modern Smart Cities(Figure 1). These silos often reflect theorganizational structures of local govern-ments. For example, the separation oflaw enforcement, transportation andpublic works into three separate depart-ments results in different IoT deploy-ments and the creation of three associ-ated technical silos.

Semantic Interoperability Solutionsfor the IoTThe interoperability of IoT systemsextends beyond technical or syntacticinteroperability, towards semantic inter-operability which mainly explores theuse of a common ontology fordescribing resources across disjoint IoTsystems [1]. Recently, several researchinitiatives have extended the ontologyof the W3C SSN incubator group,which aims to overcome the limitationsof pre-existing XML-based formats [2]and the fragmentation of sensor ontolo-gies into specific domains or applica-tions [3]. This ontology describes sen-sors, observations and related concepts,but not domain concepts (e.g., for SmartCities).

An example of a system which providesa common semantic layer is theOpenIoT project (see link below). Thissystem enables humans, devices andservices to announce and annotate dif-ferent (virtual) sensors/devices as W3CSSN compliant sensors. OpenIoT offersan open source cloud-based IoT plat-form which includes components likesensor middleware, cloud data storageand a scheduler (see link below).



Moving towards Interoperable

Internet-of-things Deployments in Smart Cities

by Gregor Schiele, John Soldatos and Nathalie Mitton

The objectives of the European FP7 VITAL project are to overcome Internet-of-Things (IoT) silos in

Smart Cities and enable the deployment and integration of Internet-Connected-Objects (ICO)

independently of the underlying IoT architecture. In achieving these goals, the project assists in the

transition towards smart, secure and cost-effective cities.

Figure 1: IoT application silos within modern Smart Cities

Figure 2: VITAL architecture

VITAL: Semantic Smart CityInteroperability The VITAL project builds on theseapproaches and extends two mainaspects to ensure the semantic interop-erability of evolving Smart City IoTapplications and projects. It uses theSSN ontology to model data and theOpenIoT as a common data manage-ment component. The first extensionprovides a much richer data model forSmart City applications, including city-wide information (e.g., demographicsand stakeholder details) as well as city-specific Key Performance Indicators.The second extension provides interop-erable access, not only for data comingfrom different IoT systems (e.g.,OpenIoT) but also for services providedby these systems (e.g., discovery, moni-toring and complex event processing).This allows for higher level services tobe created which can then be integratedinto a single federated service view.

VITAL will be tested in two SmartCities, London and Istanbul, who arerepresented in the project consortium bythe London Borough of Camden and the

Istanbul Metropolitan Municipality,respectively.

ConclusionVITAL will enable applications andservice providers to integrate servicesand data streams stemming from multipleIoT ecosystems, architectures and mid-dleware infrastructures. This will allowfor existing sensors and IoT systems to bereused and repurposed, increasing theROI on Smart City infrastructures. Wehope this will reduce the costs associatedwith developing new Smart City applica-tions for city authorities and the opendeveloper community and bring timeefficiencies, leading to a new wave ofapplications in cities across Europe.

VITAL is a three year joint Europeanproject, started in September 2013 by aconsortium of ten partners from Ireland,France, Greece, Italy, Spain, the UKand Turkey.

Links:

http://www.vital-iot.comhttp://openiot.eu/http://github.com/OpenIotOrg/openiot

References:

[1] M. Presser, P. Barnaghi, M. Eurich,C. Villalonga: “The SENSEI project:Integrating the physical world with thedigital world of the network of thefuture”, IEEE CommunicationsMagazine, pp. 1-4, 2009[2] A. Sheth, C. Henson, S. Sahoo:“Semantic Sensor Web”, IEEE InternetComputing, 12 (4), 2008[3] M. Compton, P. Barnaghi, L.Bermudez, R. G. Castro, O. Corcho, S.Cox, et. al.: “The SSN Ontology of theSemantic Sensor Networks IncubatorGroup”, Journal of Web Semantics:Science, Services and Agents on theWorld Wide Web, ISSN 1570-8268,Elsevier, 2012.

Please contact:

Gregor Schiele, NUIG, IrelandE-mail: [email protected]

John SoldatosAthens Information Technology, GreeceE-mail: [email protected]

Nathalie Mitton, Inria, FranceE-mail: [email protected]


Realizing Smart City Scenarios

with the ALMANAC and DIMMER Platforms

by Mark Vinkovits, Marco Jahn and René Reiners

Information and communication technology (ICT) is becoming a key factor to develop green and

sustainable applications within Smart City scenarios. Effective management of resources, gathering

and interpreting data as well as ecological considerations are prerequisites for realizing the vision

of smart cities. The two European FP7 projects ALMANAC and DIMMER address these issues by

providing a generic, flexible and quickly customizable platform for application development.

A Smart City service Platform (SCP)for application design and implemen-tation needs to collect, aggregate, andanalyse real-time or near real-timedata from, e.g., appliances, sensorsand actuators or smart meters thathave been deployed to implementprocesses running over a pervasivedata communication network. Theplatform must allow decision support,and implement an intelligent controlof devices through an M2M networkas well as the management of localinstal lat ions. The FP7 projectALMANAC combines real-life valida-t ions with simulat ion techniquesleading to large-scale hybrid simula-t ions that al low to test the entire

system and simulate critical loadsconditions within large scale deploy-ments.

ALMANAC focuses on the develop-ment of integrated cross-applicationinformation systems. A service deliveryplatform with corresponding tech-nology solutions integrates Internet ofThings (IoT) edge networks with com-munication access for green and sus-tainable applications. The developedservices contain management of publicand private network access - containingTelco M2M infrastructure, commonabstraction of resources, handling ofgathering latest and historical values ofheterogeneous information sources, and

reuse of data models provided by indi-vidual applications. The ALMANACarchitecture has been designed toemphasize these services, which arethen demonstrated through threeexample applications: Waste manage-ment, water management, and citizenengagement.

The four main services for the SCP thatwere indicated as initial starting pointscover interoperability over devices,such that applications can make use ofany protocol over a uniform web-service based interface. Service virtu-ality makes it possible to overcomephysical network boundaries and tosimplify application design. Composing

rules and data caches help to definethresholds and detect trends when con-crete values are not needed or theirretrieval takes too long for the currentuse case. Privacy policies of individualproviders enforce applications to onlyaccess data and functionalities forwhich they have explicit access rights.

Another aspect of smart city scenariosis covered by smart grids where “pro-sumers“ that are both, energy producersand consumers at the same time must beintegrated into the grid infrastructure.This flexibility pose new challenges tothe ICT infrastructure for planning andmanaging smart cities. The FP7European Project DIMMER (DistrictInformation Modeling andManagement for Energy Reduction) -addresses this emerging aspect thatplays a major role in the field of renew-able and independently generatedenergy. DIMMER leverages real-timedata from heterogeneous data sourcesand subsystems such as smart metersand wireless sensor networks to modelthe environmental characteristics andenergy consumption of districts. At thedistrict level this information is used tooptimize district heating and coolingand the energy grid, acting as an enablerfor the visions of smart cities and thesmart grid. The project involves rele-vant stakeholders as central players inthe research and development process,such as distribution system operators(DSOs), energy service companies(ESCOs), residents, building managers,and city planners. End user applicationsare developed targeting the needs of thedifferent stakeholders and aiming atimproving the mutual relationshipbetween them, taking into account thefact that the concept of a smart city doesnot only have a technological dimen-sion but must also consider people andinstitutions [1].

Through advanced visualization andsimulation, DSOs will be able to betterunderstand the usage of their networksand take informed decisions e.g. toimprove peak demand management andreduction, energy and cost-analysis,tariff planning and evaluation, failureidentification and maintenance.Applications for residents and buildingmanagers will help to create bi-direc-tional feedback channels betweenenergy providers, consumers and pro-sumers allowing more effective controlof the energy distribution network. In

conclusion, the expected results are aconsistent reduction in both energy con-sumption and CO2 emissions byenabling efficient energy distributionpolicies, according to the real character-istics of district buildings and inhabi-tants as well as efficient utilization andmaintenance of the energy distributionnetwork.

Figure 1 shows the described mainpillar of the project, namely stakeholderinvolvement, technology platform andbuilding/district information modelling. In order to validate the platform, bothpublic and private buildings included inmixed urban districts are considered intwo different cities, i.e., Turin (IT) andManchester (UK).

Fraunhofer FIT is responsible for thedesign and development of the softwareplatforms’ architecture, device integra-tion as well as leading the requirementsengineering process. FIT will bring intothe project its extensive knowledge andexperience in middleware development.It will employ and further develop theLinkSmart Middleware to allow inte-gration of heterogeneous technologiesand systems in the smart city districts[2].

LinkSmart, itself was established as anEU research project and is continuouslyreused and further developed in follow-up research.

Links:

http://www.fit.fraunhofer.de/en/fb/ucc/projects/almanac.htmlhttp://www.fit.fraunhofer.de/en/fb/ucc/projects/dimmer.htmlhttp://www.almanac-project.eu/news.phphttp://dimmer.polito.it

References:

[1] T. Nam, T. A. Pardo:“Conceptualizing smart city withdimensions of technology, people, andinstitutions”, in proc. of dg.o ‘11,ACM, New York, NY, USA, 282-291,2011, DOI=10.1145/2037556.2037602http://doi.acm.org/10.1145/2037556.2037602 [2] M. Eisenhauer, P. Rosengren, P.Antolin: “Hydra: A developmentplatform for integrating wirelessdevices and sensors into ambientintelligence systems”, in The Internetof Things (pp. 367-373). Springer NewYork, 2010.

Please contact:

Mark Vinkovits, Marco Jahn, RenéReinersFraunhofer FIT, GermanyE-mail: [email protected]@[email protected]



Figure 1: The three pillars of the DIMMER Project: stakeholders, the technology platform

and the BIM/DIM.

http://www.fit.fraunhofer.de/en/fb/ucc/projects/almanac.html

http://www.fit.fraunhofer.de/en/fb/ucc/projects/dimmer.html


The European project Web of ObjectsWoO (ITEA2 - 10028, January 2012-December 2014) aims to develop aservice infrastructure for IoT businessapplications that enables multi-tenancy,interoperability, and service composi-tion, thus facilitating the building ofapplications on top of embeddeddevices. The infrastructure inheritsmany features from the Web, includingbookmarking, caching, linking,searching and securing, and enablesmashups for dynamic and ad-hoc com-posite applications involving embeddeddevices. Figure 1 describes the WoOarchitecture based on IoT-A standard.

One of the technical challenges we facedon the project was creating socialobjects with web capabilities. Such fea-tures enable to have complex user-cre-ated domain oriented applications thatcan be used both in indoor and outdoorcontexts.

At home, several sensors, actuators, andappliances are available (through homeautomation technologies). In thisproject, all these resources and capabili-ties were mapped as services, and pro-vided by means of either Web ServicesTechnologies or REST Technologies [1].In addition, we used a mobile deviceplatform that provides access to dif-ferent sensors and objects (GPS loca-tions, NFC reader, vibrator, accelerom-eter, etc.) to enable the interactionbetween home objects (NFC door, homeappliances, and other sensors managedby a WSAN [2]) and mobile device sen-sors (NFC reader/writer, GPS locations,etc.). Table 1 presents an overview of thedifferent smart objects that weredeployed on the WoO platform.

In order to semantically define the smartobjects, we used and extended the SSNontology (for sensors and actuators) pro-posed in [3] and the OWL-S (for serv-

ices) to define a WoO ontology. SSNontology was also extended to supportthe definition of the “actuator” concept.By using those semantically describedsmart objects, it is possible to facilitatethe integration between smart objectsinside the WoO architecture.

On top of the WoO architecture, weprovided a service layer implementedas REST services. For instance, a doorlock web accessible service wasdesigned. This door lock can be discov-ered by means of an NFC Tag and theNFC interaction is performed through

Internet of things Applications

for Neighbourhood Embedded Devices

by Joan Fons, Daniel Gaston, Christophe Joubert and Miguel Montesinos

Smart Cities enable the physical and virtual worlds to merge, based on Internet Of Things (IoT), data

and services. Through Web of Things (WoT), IoT is realized by using the existing web architecture as

a platform, thus making smart things directly accessible as web services on the Internet. This

approach simplifies object and application deployment, commissioning, maintenance, operation and

service composition within both city and building infrastructures. In this work, we focus on

cooperative objects to develop an open smart neighbourhood.

Mobile device object

Makes use of multiple sensors (GPS, NFC, Accelerometer, etc.) to

measure physical quantities. It is used to identify the user in other

systems and to access the user’s profile. It supports the user with

GNSS (Global Navigation Satellite System) capability.

NFC Door objectOperates an NFC-tagged electromechanical door lock using a mobile

device (equipped with an NFC sensor).

User Profile object Stores users’ preferences to query about them.

GNSS Location object Represents the user's location.

Location Rules object

Represents the rules concerning location sensor-based services in

order to trigger the corresponding requests to the home automation

platform.

Request objectStores the URI and Payload corresponding to the request performed

by the mobile device.

Home Automation

Platform object

Handles the mobile device requests, and also monitors and manages

the home automation appliances.

Table 1: Social Objects in Open Smart Neighbourhood services

Figure 1: WoO architecture

an open-source mobile agent (GlooAndroid application) with profile-man-agement capabilities.

The main challenges that were faced inthis context were: smart object dis-covery (NFC tags), universal smartobject accessibility, mobile interfaceand sensor access, user profiling, andad-hoc “on-the-fly” interaction betweensmart objects (heterogeneity). Wedeveloped several scenarios, such as theLocation Sensor-Based Services out-side Home (see Figure 2). The mobileagent running on the device continu-ously receives the location coordinatesfrom its GNSS sensor and comparesthose locations with a user definedproximity area. Whenever the userenters this proximity area, the smart-phone automatically triggers a series ofrequests (based on user-profile andpreferences) to control smart objects(for example switching on the heatingto a certain temperature).

The results to date are very promising:we integrated the previous smart objectswith many partners in order to composean Open Smart Neighbourhoodecosystem on top of the Web of Objectsplatform. This allowed us to deploy andtest some recent research results in proj-ects related with the lnternet of Things,Machine to Machine communication(M2M) and Ambient Intelligence sys-tems development. We have acquiredinsight into the use of smart WoOobjects such as mobile devices, user

profiles, location rules, requests, door-locks and smart home objects.

In our research, we collaborated with sev-eral SMEs, Universities and ResearchCentres, such as DEIMOS, VISUALTOOLS, TELESPAZIO, ETIC, UPC,UPM, UPV (Spain), Thales, Odonata,Sogeti, UPEM, CEA, IMT (France),Concordia University (Canada), KAIST,KT, ETRI, Kwangwoon and HankukUniversity, Innopia, Miksistem (Korea) andSmartec, Nma, University of Cairo (Egypt).

This research is also part of a horizontaltask force with other ICT FutureInternet projects - such as FIWARE,BUTLER and SOFIA - that deals withbuilding new innovative applicationsand services for every-day working andliving environments. Our work is par-tially supported by the Spanish MECINNCORPORA-PTQ 2011, MiTYCTSI-020400-2011-29, and FEDER pro-grams.

Links:

http://www.web-of-objects.comhttp://www.youtube.com/user/webofobjectsproject

References:

[1] F. Belqasmi, R. Glitho, C. Fu:“RESTful web services for serviceprovisioning in next-generationnetworks: a survey”, CommunicationsMagazine, IEEE , vol.49, no.12, pp.66-73, 2011[2] D. Niyato, L. Xiao, L, P. Wang:“Machine-to-machine communicationsfor home energy management systemin smart grid”, IEEE CommunicationsMagazine, pp. 53-39, vol 49, 2011[3] L. Ni, Y. Zhu, and M. Jian:“Semantic Sensor Net: An ExtensibleFramework”, Springer, 2005.

Please contact:

Christophe JoubertProdevelop, Spain E-mail: [email protected]



Figure 2: GNSS

location-based

automatic house

entities requests

Internet of things:

A Challenge for Software Engineering

by Charles Consel and Milan Kabac

The Internet of Things (IoT) has become a reality with the emergence of Smart Cities, populated with

large amounts of smart objects which are used to deliver a range of citizen services (e.g., security,

well being, etc.) The IoT paradigm relies on the pervasive presence of smart objects or “things”,

which raises a number of new challenges in the software engineering domain.

The Object’s World projectThere are an abundance of research andindustry initiatives that have been under-taken with the aim of promoting theemergence of IoT [1]. In line with thisgoal, the Object’s World project bringstogether stakeholders from differentdomains to build and support the emer-gence of an IoT sector in France and

beyond. The project is lead by SIGFOX,the world's first cellular network oper-ator dedicated to low-bandwidth wire-less objects. The cooperation betweenindustry and research partners (e.g.,sensor manufacturers, computer scienceand electrical engineering research labs)is of uttermost importance in over-coming technological barriers. This issue

is currently hindering the developmentof an IoT sector. The main objectives ofthis project are the development of:• expertise in the low-bandwidth net-

work sector,• low-cost transmitter/receiver chips,• low-energy autonomous sensors, and• software frameworks which cover the

entire lifecycle of IoT applications.

http://www.youtube.com/user/webofobjectsproject


Network infrastructures which supporthuge numbers of objects open up a rangeof opportunities for innovative services.Critically, these new opportunities relyon the ability to address the softwareengineering challenges of this newsector. We promote an approach thatrevolves around software frameworks.In areas such as mobile and web devel-opment, this approach has already beenshown to facilitate software develop-ment by abstracting over implementa-tion details and guiding the programmer.

A design-driven development approach To guide and support the developmentof applications which orchestrate net-worked objects, our research groupintroduced a design-driven softwaredevelopment approach which draws onprinciples and techniques taken fromthe programming language domain.

In particular, we developed DiaSuite [2],a tool-based methodology which guidesthe developer through the entire life-cycle of an orchestrating application(Figure 1). DiaSuite offers a design lan-guage, providing high-level, declarativeconstructs that are dedicated todescribing the application’s architecture,along with the smart objects it orches-trates. The methodology relies on a com-piler that generates support in the formof a Java programming framework, cus-tomized with respect to a given applica-tion design. By providing the developerwith a programming framework, ourapproach ensures conformity betweenthe design and implementation. Thegenerated programming framework pro-vides the developer with an abstractclass per component declaration. Theapplication logic is implemented by sub-classing each abstract class and pro-gramming its abstract methods. To fur-ther ease the development of orches-trating applications, DiaSuite relies onEclipse to guide developers during theimplementation phase by introducingplaceholders that need to be providedwith code. Finally, our approach pro-vides developers with a back-end toaddress the deployment and executionof orchestrating applications.

Orchestrating smart objects at a large scaleThe development of orchestrating appli-cations which are responsible for largenumbers of smart objects raises a numberof challenges. We have addressed theseby introducing a new design language.

Service discovery

Standard service discovery at the indi-vidual object level does not address theneeds of applications orchestratinglarge numbers of smart objects. Instead,a high-level approach which providesconstructs to specifying sub-sets ofinterest is needed. Our approach allowsdevelopers to introduce application-specific concepts (e.g., regroupingparking spaces into lots or districts) atthe design time and then these can be

used to express discovery operations.Following our design-driven develop-ment approach, these concepts are usedto generate code to support and guidethe programming phase.

Data gathering

Applications need to acquire data froma large number of objects through avariety of delivery models. Forinstance, air pollution sensors across acity may only push data to the relevantapplications when pollution levelsexceed tolerated levels. Tracking sen-sors, however, might determine thelocation of vehicles and send theacquired measurements to applicationsperiodically (e.g., 10 min. intervals).Data delivery models need to be intro-duced at design time since they have adirect impact on the application’s pro-gram structure. In doing so, the deliverymodels used by an application can bechecked against sensor features early inthe development process.

Data processing

Data that is generated from hundreds ofthousands of objects and accumulatedover a period of time calls for efficientprocessing strategies to ensure therequired performance is attained. Ourapproach allows for an efficient imple-

mentation of the data processing stageby providing the developer with aframework based on the MapReduce [3]programming model which is intendedfor the processing of large data sets.

Future workWe envisage to enrich our design-driven methodology with support forsimulation of infrastructures of smartobjects. To achieve this, we willleverage design-time declarations to

generate application-specific simulationsupport, while keeping the applicationcode unchanged.

Links:

http://www.sigfox.com/en/http://www.telecom-design.com/en/http://phoenix.inria.fr/

References:

[1] D. Miorandi et al.: “Internet ofthings: Vision, applications andresearch challenges,” Ad HocNetworks, vol. 10, no. 7, 2012[2] D. Cassou, et al., “Towards A Tool-Based Development Methodology forPervasive Computing Applications,”IEEE Transactions on SoftwareEngineering, 2011[3] J. Dean and S. Ghemawat,“MapReduce: simplified dataprocessing on large clusters,”Communications of the ACM, 2008

Please contact:

Charles ConselUniversity of Bordeaux / InriaBordeaux - Sud-Ouest,E-mail: [email protected]

Milan KabacInria Bordeaux - Sud-Ouest,E-mail: [email protected]

Figure 1: The DiaSuite tool-based methodology

The different types of moving objects ina city, such as people, vehicles andassets, that are equipped with mobiledevices (e.g., smartphones, wearabledevices, smart sensors, etc.) haveincreasing computing, communicationand sensing capabilities. Thus, they willplay a key role in Smart Cities. The effi-cient management of the informationgenerated, provided and used by theseobjects (including information on theirlocations, trajectories, features, behav-iors, activities and environmental con-texts) facilitates the improved under-

standing and analysis of how a city per-forms. In addition, this type of informa-tion assists in the goal of providing citi-zens with contextual and adapted serv-ices in a range of areas including trafficmanagement, urban dynamics analysis,ambient assisted living, emergencymanagement, and mobile health. Whilstsignificant effort has been invested inmodelling and managing movingobjects, and some progress has beenmade regarding the representation ofsemantics associated with them, furtherefforts are needed to achieve a fully-

fledged semantic managementapproach for moving objects that can beefficiently and flexibly exploited.

By integrating methods and techniquesdeveloped in different fields, includingmoving object databases, spatio-tem-poral data mining, and the SemanticWeb, it is possible to enhance the waymoving object information is managed(with respect to the modelling,querying, processing and analysis com-ponents). For example, recent proposalsclaim that linking Location-Based



Semantic Management of Moving Objects

in Smart Cities

by Sergio Ilarri, Dragan Stojanovic and Cyril Ray

Smart Cities depend on information regarding moving objects (e.g., people, vehicles, assets, etc.)

being processed. We propose a framework to enable a fully-fledged semantic management of

moving objects that can be efficiently and flexibly exploited in Smart Cities.

Figure 1: The architecture of the SemanticMOVE framework

Services (LBSs) and semantics can pro-vide interesting benefits [1], such asflexible querying, management ofsemantic locations and trajectories,interoperability among different LBSsand providers, protection of personallocation information, and reasoning incomplex and dynamic contexts. Ouridea is to develop and provide LBSs thatunderstand the user’s requests and inter-actions, implicitly based on the seman-tics of mobility and contextual informa-tion, and know how to behave and adaptin dynamic and unexpected situations.

Incorporating semantics in such anInternet of (Moving) Things providesvaluable semantic information andservices to mobile users, thus sup-porting enhanced mobility in dynamicenvironments such as those found inSmart Cities. However, there arenumerous research challenges that needto be adressed to make this idea areality. The first lies in the collaborativecollection of the data required (e.g., thedata measured by sensors on themoving objects). The data collectioncan be achieved through participatory/collaborative sensing but this is subjectto difficulties related to the correlationand analysis of those data, especiallywhen this analysis is performed in a dis-tributed way on mobile devices.Another challenge relates to thesemantic representation of movingobjects, where a unified approach thattakes all the mobility aspects of movingobjects (i.e., their trajectories, contexts,activities, goals, the physiologicalstatus of the user, environment, and theaccessible services) into account, at dif-ferent levels of granularity, is stillmissing. Thirdly, while interesting workhas progressed on context-awareness,there is still a need for efficient inte-grated approaches for query processingin mobile environments, reasoning, andsemantic searches, as well as an appro-priate abstraction layer that enables theexploitation of the available functional-ities. The fourth challenge relates to theinference of higher-level semantics(such as group mobility behaviours)from a large dataset of individualsemantic mobility data and trajectories.Finally, we have to mention the problemof privacy protection. This is a particu-larly critical issue in Smart City envi-ronments, as the basic movement data isenhanced with rich semantics of theparticipating moving objects and their

Links:

MOVE: Knowledge Discovery fromMoving Objects:http://dmsl.cs.ucy.ac.cy/projects/cost/index.html SemanticMOVE:http://webdiis.unizar.es/~silarri/SemanticMOVE/

References:

[1] S. Ilarri et al.: “Semantics inLocation-Based Services - GuestEditors’ Introduction for SpecialIssue”, IEEE Internet Computing,ISSN 1089-7801, 15(6): 10-14, IEEEComputer Society, 2011 [2] B. Predic ,D. Stojanovic:“Localized Processing and Analysis ofAccelerometer Data in DetectingTraffic Events and Driver Behaviour”,Journal of Universal ComputerScience, ISSN 0948-695x, 18(9):1152-1176, 2012[3] I. Afyouni, C. Ray, C. Claramunt:“Spatial Models for Context-AwareIndoor Navigation Systems: ASurvey”, Journal of SpatialInformation Science, ISSN, 4(1): 85-123, 2012.

Please contact:

Sergio IlarriUniversity of Zaragoza, SpainE-mail: [email protected]

Dragan StojanovicUniversity of Nis, SerbiaE-mail:[email protected]

Cyril RayInstitut de Recherche de l’ÉcoleNavale (IRENav), FranceE-mail: [email protected]


trajectories. Thus, it should be pre-served according to the required privacypreferences of the participating users.

Prompted by the COST Action IC0903on knowledge discovery from movingobjects, we have collaborativelydesigned a generic and scalable distrib-uted framework (SemanticMOVE),whose realization will enable the com-prehensive management of the seman-tics of moving objects. Achieving thiswill leverage increased sensing, pro-cessing, interaction and communicationcapabilities in mobile devices in a scal-able and effective way (Figure 1). Asopposed to other related work, we envi-sioned a quite generic architecturewhich supports a fully distributed andinteroperable scenario for the manage-ment of semantics of moving objects.Thus, each moving object can collect,store and analyze its own semanticmobility data and trajectories, andreason over them locally. It can shareand exchange semantic mobility dataand semantic concepts/knowledge (e.g.,regarding the location, personal andsocial status, vehicle conditions, activi-ties, behaviours, environment andtraffic conditions, air pollution levels,etc.), with other movingobjects/users/services in the vicinity,over ad-hoc wireless networks. It canalso access and share information andknowledge through geo-social net-works, social media services, andgeospatial information services. Ourframework includes fixed servers as anadditional element of an ecosystemwhere the distributed and ad-hoc coop-eration among moving objects, encour-aged by some incentive mechanisms,plays a key role.

We believe that the development ofsuch an approach will facilitate thedeployment of interesting semanticmobility applications for Smart Cities.Whilst there are still some significantresearch issues that require furtherinvestigation, the initial results repre-sent promising steps towards theintended direction [2, 3].

http://dmsl.cs.ucy.ac.cy/projects/cost/index.html

http://webdiis.unizar.es/~silarri/SemanticMOVE/

As Smart Cities become a reality, citi-zens are starting to be overwhelmedwith the amount of data they receivefrom different sources. This is partlycaused by the sheer number of apps theycan download to obtain information:most apps are designed for specific sce-narios and goals and are embedded withimplicit knowledge about the applica-tion context. They also receive dataalmost continuously (e.g., informationabout pollution, traffic, parking spots,lighting, etc.), which makes it difficult todistinguish which information is valu-able. In this scenario, the use ofsemantic techniques becomes particu-larly relevant: a system that usessemantic information can help usersselect the most appropriate and relevantapps to the user's interests and transformthe raw received data into smart datathat represents actionable information.

Mobile devices (e.g., smartphones andtablets) have become a fundamental partof our everyday lives. According to areport by BI (Business Intelligence), onein five people worldwide owns a smart-phone and one in 17 owns a tablet. Thesedevices not only consume informationbut also create huge volumes of data(e.g., geo-tagged images, videos, text,etc.). In addition, sensors are ubiquitousand a common feature on smart andwearable devices (e.g., activity trackers,smart glasses and watches) as well as

city-based sensors (e.g., sensors thatmeasure air pollution, noise and trafficlevels). These city-based sensors areuseful in the context of Smart Cities.New management approaches arerequired for all this information toensure users do not become over-whelmed. The semantic management ofdata in wireless environments can helpusers in a variety of ways includingassisting with the process of deter-mining what information a user reallyneeds to finding the most appropriateinformation from a range of differentsources and presenting it in an integratedway. For example, imagine tourists whoarrive on an evening flight and need toreach their city hotel. At first, it wouldbe useful to infer what kind of informa-tion the tourists might need, forexample, with regards to transport infor-mation they might need to know the dif-ferent options (e.g., buses, metros, taxisor car rental options), traffic conditionsand perhaps even where availableparking spaces are located. This infor-mation comes from a variety of hetero-geneous sources (e.g., websites, city-based sensors, other users and vehicles)and it is difficult to reconcile and presentthese data as they are needed.

In developing the SHERLOCK [1]system, we sought to provide mobileusers with interesting Location-BasedServices (LBSs). Begun in 2011, this

collaborative project brings together theDistributed Information Systems (SID)group at the University of Zaragoza(Spain) and the Interoperable DatabaseGroup (BDI) group at the BasqueCountry University (Spain). Since westarted, we have developed an Androidprototype [2] that can be downloadedfrom the website of the project (seeFigure 2 for a screenshot of the app).

Our main focus is to develop a generaland flexible system that is able to respondto the information needs of mobile users.We met a number of challenges in thisproject, but there were three main prob-lems. Firstly, the system has to be able tounderstand the information needs of theuser. For this task, SHERLOCK lever-ages the user’s context (e.g., location,activity, etc.) that can be extracted for arange of sources, for example, from sen-sors in his/her mobile device. A personalagent guides the user in the process ofselecting an interesting LBS. This agentuses context information along with



flexible Access to Services in Smart Cities:

Let SHERLOCK Advise Modern Citizens

by Roberto Yus, Eduardo Mena, Sergio Ilarri and Arantza Illarramendi

Citizens can access a variety of computing services to get information, but it is often difficult to

know which service will offer the best information. Researchers in the SHERLOCK (System for

Heterogeneous mobilE Requests by Leveraging Ontological and Contextual Knowledge) project,

from the University of Zaragoza and the Basque Country University, address this by providing

mobile users with interesting Location-Based Services (LBSs).

Figure 2: SHERLOCK app for the transports

scenarioFigure 1: SHERLOCK finding appropriate transport in a city

information about the services, encodedin ontologies (i.e., a formal representationof knowledge) and a semantic reasoner(i.e., a software to infer logical conse-quences), to deduce which LBSs wouldbe useful for the user. Secondly, thesystem has to be able to find the informa-tion the user needs. For this task, SHER-LOCK deploys a network of mobileagents (i.e., software that is able to movefrom one device to anotherautonomously) to search all availableinformation among the distributed datasources and bring back the results tohis/her device. Finally, the system has tobe able to present the results to the userand keep this information updated for aslong as the user requires.

We believe that a system such as SHER-LOCK is very useful in the context ofSmart Cities, as it can support citizensand visitors in achieving their tasksmore efficiently and effectively. Anydeveloped city services could easily bemade available to SHERLOCK, simplyby defining it in the form of anontology.

Links:

SHERLOCK:http://sid.cps.unizar.es/SHERLOCKSmartphone and tablet penetration in2013: http://www.businessinsider.com/smartphone-and-tablet-penetration-2013-10

References:

[1] R. Yus, et al.: “SHERLOCK:Semantic Management of Location-Based Services in WirelessEnvironments”, Pervasive and MobileComputing, 2013.[2] R. Yus, et al.: “SHERLOCK: ASystem for Location-Based Services inWireless Environments UsingSemantics”, WWW 2013

Please contact:

Roberto YusUniversity of Zaragoza, SpainTel: +34 976 76 26 50E-mail: [email protected]


Quantifying the Benefits of taxi trips

in New York through Shareability Networks

by Paolo Santi, Giovanni Resta and Carlo Ratti

Shareability networks demonstrate that more than 95% of taxi trips taken in New York can be

shared with minimal passenger discomfort.

The increasing pervasiveness of digital-ized information has unleashed unprece-dented opportunities for understandingaspects of human behaviours and sociallives, including individual mobility. Anenormous amount of digital traces arenow obtainable from a range of sources(e.g., cell phone records, taxi GPStraces, etc.) which allows humanmobility to be analyzed to an extent thatwould have been inconceivable severalyears ago [1]. Although this raises pri-vacy concerns, this “big data” era offersunique opportunities to improve under-standing around human mobility needsand, hence, improve transportationsystem efficiencies.

The goal of this project, performed incooperation with a team from the MITSenseable City Lab (Michael Szell, StanSobolevsky and coordinated by CarloRatti), is to quantify the benefits of taxisharing in New York City (NYC). Ouranalysis was based on a dataset whichcaptured all the taxi trips taken in NYCin 2011 (over 150 million trips). Foreach trip, the dataset captures the pick-up time and location and drop-off timeand location.

In this project, we posed the funda-mental question, “How many taxi trips

can be shared in NYC?”. To answer thisquestion, the intrinsic trade-off betweenshareability opportunities and pas-senger discomfort must be considered:the longer a passenger is willing to waitfor a shared trip, the higher the sharingopportunities. This tradeoff is madeexplicit by the novel notion of a share-ability network in which we havedefined the model sharing opportuni-ties: each network node represent a sep-

arate trip and links between two nodesrepresents a sharing opportunitybetween those trips. The criterion usedto determine whether two trips can beshared is based on spatial and temporalconstraints. For two trips, T1 and T2, asharing opportunity only exists if aroute connects the respective pick-upand drop-off points such that both pas-senger groups can be picked up anddelivered to their destinations with a

Figure 1: The shareability of taxi rides taken in New York City is constrained by the length of

time customers are willing to be delayed. As the delay time lengthens, trip sharing opportunities

increase. Results indicated that a delay of 5 min meant more than 95% of the taxi rides could be

shared.

http://www.businessinsider.com/smartphone-and-tablet-penetration-2013-10

delay of no more than ,where explic-itly models the tradeoff between pas-senger discomfort and shareability. Ahigher value of results in a densershareability network and corresponds tomore opportunities for trip sharing.

Using a shareability network allows anoptimal solution to be found to an other-wise computationally intractableproblem. Shareability networks imposea structure to an otherwise unstructured,immense search space by constrainingthe number of trips that can be shared(up to k, where k is a user-definedparameter, set to 2 or 3 in this study) andconsidering only static trip sharing. Thisterm means that once two or more tripsare combined into a shared trip, thecombined trip is served by a single taxithat cannot be rerouted for furthersharing. Once the search space has beenreduced and structured, an optimal tripsharing figure can be computed in

approximately 0.1 seconds by running acomputationally efficient maximummatching algorithm on the shareabilitynetwork (10,000 nodes and 100,000links) [2] using a standard Linux work-station. Thus, our proposed method-ology is suitable for real-time imple-mentation.

A notable result of our analysis is thatconstraining the search space to reducecomputational complexity does notimpair trip sharing opportunities. Thepercentage of shareable trips is shownto increase with the delay parameter andwe found it was as high as 95% if adelay in the order of five minutes waspermitted. Thus, the vast majority oftaxi trips taken in NYC can be sharedwith minimal passenger discomfort.

We also investigated the effects of thesharing penetration rate which accountsfor the true fraction of passengers that

want to use a shared taxi service. We areextending this analysis to other cities(Singapore, Vienna, San Francisco, etc.)to investigate whether similar sharingopportunities arise in other urban con-texts.

References:

[1] M. Gonzales, C. Hidalgo, A.L.Barabasi, “Understanding IndividualHuman Mobility Patterns”, Nature,2008[2] Z. Galil, “Efficient Algorithms forFinding Maximum Matching inGraphs”, ACM Comp. Surv. 1986.

Please contact:

Paolo SantiIstituto di Informatica e Telematica andMIT Senseable City LabE-mail: [email protected];[email protected]



Integrated Electric Vehicles Sharing and Pooling

Mobility Solutions for Smart Cities

by Marie-Laure Watrinet, Gérald Arnould, Hedi Ayed and Djamel Khadraoui

Integrating public transport systems with individual car-and-ride sharing concepts is considered as an

attractive, convenient and emissions reducing mobility concept in the frame of Smart Cities. The

pooling of mobility services is considered to be an important enabler of the Smart Cities’ concept,

especially with regards to achieving flexibility and integration with existing transport modes (mostly

public transport). Even if the levels of user acceptance towards new smart sustainability concepts are

still challenging their uptake, it is important to address ICT-related challenges to ensure adaptive

solutions are found to reduce complexities. This is especially relevant in the case of electro-mobility

related systems.

Tudor, via its mobility projects, hasdeveloped a concept for sharing electricvehicles (EVs) and cross-company opti-mization solutions. This concept aims toincrease sustainable resource produc-tivity by sharing EVs across variouscompanies or organizations. Thisapproach is likely to increase the overallusage of each vehicle but reduce costsper kilometre and users. During thework-day the vehicles are used for pro-fessional activities, but for the rest of thetime, they become a collective resourcethat can be placed in carpooling mode.

Such a combined usage presents signifi-cant algorithmic complexity and solvingthis so that an e-fleet can be used forpublic and professional purposes is chal-lenging, The actual concept was evalu-

ated via simulations of different sce-narios using MATSIM, A multi-agentapproach was considered, where eachagent represented a traveler. A previousnational project, Moebius [1,2], pro-vided the simulation data. The mainobjectives were to validate the conceptof combined car-sharing and car-pooling from the perspective ofresource optimization and find strategiccharging locations [3].

The EV sharing service concept pres-ents users with a hop-on, hop-offsystem that has demand responsivefleet management (which includes pre-dictions of what locations are going tohave the highest user needs). Thesystem’s resources (e.g., EVs, e-bikesand associated infrastructures) are

mutualized along with an informationsystem, the service and, in future,public transport services.

The goal of the mutualisation conceptis to maximize the use of vehiclesacross the day (both work and recre-ational times) but decrease the residualcost of the EVs. This can be achievedwith an optimization algorithm [1] thatcan minimize the required number ofvehicles at all times, based on plannedusage (determined using statistics andsimulation tools), real-time demand,and third party companies that canmanage the fleets (by zones) with newbusiness models. EV sharing can alsobe associated with other transportmeans such as car-pooling and publictransport.


The experimentationsTwo real experimentations (via twoEuropean Regional Development Fund(ERDF) projects called ZAC-eMovinand Nordstad-eMovin) have beendeployed under the actual concept. Thefirst project is on the professional use ofshared EV fleets and infrastructurebetween companies. The secondfocuses on the citizen oriented integra-tion of several types of e-mobility serv-ices. In both projects, a coherentapproach is considered, including theuse of public transportation systems andthe identification and integration of therequired IT services. Each project pro-vide recommendations on the best wayto integrate e-mobility services, basedon in-depth analyses of relevant sys-tems and user behaviours.

In the ZAC-eMovin project, employeesof the three project partners are usingthe EV in their company’s fleet for day-to-day professional travel as well as

travels between home and work (andvisa versa). The goal of the project is tomutualize the resources, EV andcharging infrastructure in an activityzone between the companies.

The fleet management service relies ona resource allocation algorithm whichtake the limited range of EVs and theircharging periods into account, fol-lowing which it optimizes the EV. Theservice proposes public transport tripsinstead of EV leases where possible,and allows for resources to be sharedacross the participants which guaran-teeing safe data exchange and privacy.

The Nordstad-eMovin project is consid-ering the shared usage of EVs and e-bike in Nordstad (Luxembourg) fromthe citizen’s perspective. This project isalso considering the integration ofpublic transport systems. Citizens canuse any of the three services offered bythe system in an integrated way: EV

short-time renting, e-bike renting or pri-vate vehicle charging. E-bikes are pre-ferred for short-distance travel or whenthere are heavy traffic conditions. EVscan be used to go shopping, or forlonger-distance travel, depending on therequirements of the user. A typicalusage could cover the last mile of travelfrom the train station to work or home.Resources will be mutualized to allowfor both the professional and private useof the services.

Links:

http://www.nordstad-emovin.luhttp://www.zac-emovin.lu

References:

[1] H. Ayed, D.Khadraoui: “LocalSearch Algorithm for the Daily CarPooling Problem”, InternationalConference on Metaheuristics andNature Inspired Computing 2014[2] H. Ayed, B. Gateau, D. Khadraoui:“Simulating daily mobility inLuxembourg using multi-agenda basedsystem”, 28th European Conference onModelling and Simulation 2014.[3] G. Arnould, et al.: “Modelling andsimulating a dynamic carpoolingsystem for improving citizensmobility”, International Journal onSpace-Based and Situated Computing,Vol 1, 2013.

Please contact:

Marie-Laure Watrinet, Gérald Arnould,Hedi Ayed, Djamel KhadraouiCRP Henri Tudor, LuxembourgE-mail: [email protected],[email protected]@tudor.lu,[email protected]

Figure 1: Zac-eMovin mutualized mobility service

infrastructure

Figure 2: Nordstad-eMovin mutualized mobility services

Figure 3: Service mutualisation in three activity zones

In recent history, cars have become ourprimary choice of transport. However,this high number of motor vehicles havebecome a significant contributor to cli-mate change, emitting approximatelyfive billion tons of carbon every year.The transportation sector currentlyaccounts for approximately 15% ofoverall greenhouse gas emissions, afigure which grew by 45% between1999 and 2007. While getting drivers toswap the comforts of a car for publictransport seems like an unattainablegoal, encouraging them to carpool is amore appealing prospect and with anestimated 3.75 empty seats per car, pertrip [1], the efficiency potential is signif-icant. Changing the mindset of car com-muters might not be the task of com-puter science, but making carpoolingservices more accessible definitely is. Inthe era of smartphones, with locationtracking abilities and always-on connec-tivity, providing the public with an easy-to-use recommendation system is nolonger impossible.

According to our vision here at theBudapest University of Technology andEconomics, the ideal computer-aidedcarpooling system should require verylittle effort from the users. In the last twoyears, research engineers at theDepartment of Automation and AppliedInformatics developed the underlyingalgorithms such a system would use, andwe describe that system below.

It depends on a light-weight applicationruns constantly on the smartphones ofboth the drivers and the passengers,tracking their locations and learningtheir transport usage habits. This data isnot uploaded to any server, but rathermovement patterns are extracted on thephone itself using advanced, mobile-optimized algorithms. The recurring pat-terns, identified as routines, consist of aday reference (i.e., what days of the

week they happen on), the approximatetime window (e.g., between 7:30 and8:20) and the actual route typicallytaken by the user.

When a user wants to either volunteer asa driver or seek a ride, they can securelyupload their routines to a central server,where the matching process takes place.Here, the routines of the drivers andpassengers are compared to find good

matches which are then suggested to theuser. For example, the usual departureand destination locations of PassengerX are indicated by blue markers inFigure 1. After cross-comparing hisdeparture and destination preferenceswith the morning routines of drivers inthe same area, the system found twopotential cars Passenger X may be ableto join (displayed by the red and greentracks). Note that the full routes areintentionally not shown since they con-sidered sensitive, private information.When Passenger X receives his sug-gested options, he can clearly see theproposed pick-up and drop-off loca-tions, as well as approximate timings.He can also call the driver using the

phone icon displayed next to theirname.

This system assumes that arrival timesare strictly held on morning routes sincepeople are typically going to work atthis time. Conversely, with afternoonroutes, departure times are more impor-tant since users don’t want to wait longfor their rides after work. Therefore,when a carpooling search is performedfor the morning, the system considersthe distance between the drop-off anddestination points and (using a five kilo-meter per hour walking speed) calcu-lates the possibility of being able towalk to the destination and arrive beforethe strict arrival time.

While these carpooling recommenda-tions a very useful feature by them-selves, the large set of learned routinesenable a much broader family of serv-ices. For example, in case of a trafficjam or road works, the system is able toidentify their location(s) by identifyingthe lower car speeds of drivers who usu-ally take that route. Other regular routeusers could then be automatically noti-fied of such a disruption, thus enablingthem to leave at a different time or takea different route. Our team is currentlyworking on supporting several usagesof this geo-social information.

Links:

http://amorg.aut.bme.hu/http://www.internationaltransportforum.org/2010/

Reference:

[1] Mayinger, F. (Ed.). Mobility andTraffic in the 21st Century. Springer.

Please contact:

Marcell FeherBudapest University of Technologyand Economics, HungaryE-mail: [email protected]



A Carpooling Recommendation System

in the Smartphone Age

by Marcell Fehér and Bertalan Forstner

Worldwide, the one billion cars in use emit enormous amounts of carbon dioxide. Potentially these

emissions could be drastically reduced if people shared their daily commutes. The idea of carpooling

has been around for a long time and already, millions of gallons of gas have been saved, but the

widespread adoption of this strategy is yet to come. Here, we introduce a carpooling service which is

easy to use and aims to minimize user effort, thus encouraging the switch to this smart,

environmentally friendly transportation option.

Figure 1: An example of the proposed

carpooling system where Passenger X’s usual

departure and destination locations are

indicated with the blue markers. Following a

comparison process with drivers in the same

location, two possible carpooling suggestions

are identified (indicated by the red and green

tracks).

http://www.internationaltransportforum.org/2010/


Under the promise of improving themanagement and efficiency of cities, theconcept of Smart Cities has gained rele-vance over recent years. Local councilshave promoted individual Smart Citypractices through investments in dif-ferent sensor systems, infrastructures,measurements and modernisation. Anenormous amount of data is now avail-able, collected from sensors with dif-ferent architectures and specifications.The challenge in turning this into rele-vant information for citizens. Thisproject aimed to research and develop asingle framework for the design of anintegral Smart City strategy based on asingle plug&play scalable system anddiverse data sources.

The Quercus Software EngineeringGroup, University of Extremadura incollaboration with the Vodafone SpainFoundation designed (2013) andlaunched (2014) the first version of aSmart City living lab at the PolytechnicSchool of Cáceres, thus transitioning itinto a Smart Campus. A campus envi-ronment features the majority of behav-iours that occur across the broader cityenvironment.

One of the main problems that citizensand municipalities face is the mobility

and parking. This is the main motiva-tion for the Smart Parking project. Thegoal of this project was to design a low-cost single system that allowed for dif-ferent data sources to connect in aplug&play mode, being scalable andmainly based on software engineeringtechniques. This system would 1) offerdrivers real-time assistance to findavailable parking spaces near them, 2)know the occupation ratios for takingdecisions about parking space vehicleoccupancy and 3) provide useful infor-mation for disabled people who typi-cally require more time to find freeparking spaces.

In this context, sensors were installed inthe parking spaces of the Cácerescampus, using three different datasource technologies. We aim to com-pare reliability, costs, performance, andother factors between the different tech-nologies [1] in an empirical way todetermine which is best.

Hardware architectureThe architecture of the system is basedon a three-layer model, Producers-Server-Consumers, in which sensordevices are responsible for producing(PRODUCERS) and sending data to theserver [2]; the server (SERVER) gives

intelligence to the system using algo-rithms that transforms the data into rele-vant information; and the users (CON-SUMERS) that query this informationin real-time.

The layer of the data producers (PRO-DUCERS) is formed by three types ofsensors which are outlined in Table 1.

The data produced by the sensors aresent to the server, where the occupancystatus of each parking spaces is stored.This layer is responsible for providingintelligence to the system, processing alldata captured and transforming theminto relevant information. Then the userscan access the services offered by thesystem from any device with internetaccess. They can check the real-timeoccupancy status of parking spaces andview the history of occupation. It is alsopossible to know if a space is under- orover-utlized. The Smart Parkingarchitecture is shown in Figure 1.

Software architectureThe software layer has been developedusing web engineering technologies(WebRatio 7.1/HTML 5) and provideaccess to any user from any device withinternet access. The main view of thesystem is shown in Figure 2.

A Smart Parking Campus:

An Example of Integrating Different Parking

Sensing Solutions into a Single Scalable System

by Enrique Moguel, Miguel Ángel Preciado and Juan Carlos Preciado

Smart Parking is based on a software system that links hardware and software, with support from

Augmented Reality technologies, to provide an enhanced solution in the query of information regarding

parking spaces. Smart Parking responds to the need of people with disabilities who have to know the

availability of adapted parking spaces.

Table 1: The three types of data producers Figure 1: The hardware architecture of the Smart Parking system

Augmented reality layerIn innovative addition has been devel-opment of an augmented reality system[3] that provides the user with context toaround where the parking spaces areplaced. This differential technologysolution can help any user find the cor-rect position of a parking space and itsoccupancy status.

Conclusions and future linesDeveloping a single plug&play scalablemulti-layer system has enabled us toconceptualize the design of a Smart City(parking dimension). This developmentprocess was founded on our extensiveR&D of systems engineering but inno-vation opportunities that came from les-sons learnt also played a key role. Theselessons were from experiences in a widerange of fields including big data, cloudcomputing, business intelligence, busi-ness analytics, data visualization, opendata and packed services.

This experience has allowed us tomanage Smart City concepts at the labo-ratory level and investigate differentsoftware/hardware alternatives. Asystem based on software engineeringoffers a scalable and plug&play designand at the sensing level we can developa low-cost product.

We are also managing cameras (avail-able in Figure 3) which sense parkingplaces via software. In investigating thisavenue, we are seeking to maximise theopportunities (i.e., cheaper infrastruc-ture, more control about cars) but solvethe associated problems (i.e., moreserver processing).

Acknowledgement: Work funded by Spanish ContractMIGRARIA - TIN2011-27340 atMinisterio de Ciencia e Innovacion andGobierno de Extremadura (GR-10129)and European Regional DevelopmentFund (ERDF)

Link: http://uex.be/SmartSpaces

References:

[1] G. Revathi, V.R.S.Dhulipala: “Smart parking systems and sensors: A survey”,in proc. of Computing, Communication and Applications (ICCCA), Prag 2012,p.1-5[2] U. Männi: “Smart sensing and time of arrival based location detection inParking Management services”, in proc. of Electronics Conference (BEC) 2010, p.213-214.[3] P. Pulli: “Mobile Multimodal Virtual Reality for Personal Teleservices”,ERCIM News No.31, October 1997.

Please contact:

Enrique Moguel, University of Extremadura,, Spain.E-mail: [email protected]

Miguel Ángel, Preciado, Homeria Open Solutions, Spain. E-mail: [email protected]

Juan Carlos, PreciadoUniversity of Extremadura, Spain.E-mail: [email protected]



Figure 4: The augmented reality layer

which provides users additional spatial

context so that they can find the correct

parking space.

Table 2: Results obtained in the sensors comparison.

Figure 2: Real-time view of the occupancy status for each parking space

in the campus parking lot.

Figure 3: Distribution of devices by type and status of parking spaces.


When commuting via public trans-port, a common problem is findingthe right direction. Typically gettingto the correct destination requiresseveral bus or train changes. Manypublic transport providers, as well asindependent companies, offer onlineplanning tools to simplify this task.The use of these tools usuallyinvolves the user entering the originand destination locations for a givenjourney along with nominating sev-eral preferences (e.g., maximumnumber of interchanges). The resultsthen provide the user with asequence of directions, nominatingthe different buses and trains thatmust be taken. Ideally, multiple optionsare suggested, each with its own advan-tages and disadvantages. For example,Option 1 might present a route with ashort travel time, but requiring multiplechanges, Option 2 a slow but directroute, and Option 3 may be the cheapestoption but it includes an extended waittime at an intermediate stop. For anygiven option, missing one step in thesequence can result in the routerequiring re-planning to address the dif-ferent conditions now faced by the user.

A traveller who is familiar with a PTSmight blend aspects from the variety ofsolutions presented to develop their ownunique solution. This offers the user theadvantage of being able to adjust theirjourney dynamically in response to theirown shifting needs. For example, if theuser misses a connecting bus, he/shemay be aware of an alternative buswhich also suits their current needs. Thisdynamism is especially necessary if thePTS is unreliable and features “rule ofthumb” scheduling or only publishesvehicle frequencies as opposed to actualtimes (e.g., three buses per hour vs. threebuses at 11:20am, 11:40am and12:00pm).

Challenge and Policy-based PlanningTurning this intuitive planning approachbased on personal experiences into a sto-chastic planning tool presents an

ongoing challenge which the authorshave addressed using a number of dif-ferent approaches. Instead of providinga simple sequence of options to follow,these approaches all suggest a tree ofoptions, called a policy. The policy out-lined above (Figure 1) includes fourstops A, B, C and D: at each stop thetraveller can pick one of the providedalternatives. For example, at Stop A theuser can select Bus 1 or Metro 1.Choices can be made according to afirst-come-first-serve rule which is easyto execute (even in an offline mode) orto a more detailed timing regime whichmight be supported by a smartphoneapplication. We have developed severalalgorithms to compute such policies.

Drawing on the heuristic search used inthe Artificial Intelligence area, the firstalgorithm uses a forward planning per-spective (i.e., the user is at the begin-ning of their journey) [1] [2] andattempts to factor in all the possibleevents and options to find a policy thatis flexible within each situation. Thesecond algorithm, inspired by the clas-sical shortest path algorithms, takes abackward planning perspective usingdynamic programming techniques [3].Thus, this algorithm iteratively buildsstochastic plans starting from the desti-nation. Both the algorithms we devel-oped are fast enough to allow for real-time application.

There is, of course, the need to buildan abstraction of reality in bothalgorithms. For example, the sto-chastic behaviour of buses must besimplified to make their movementscomputationally tractable.Therefore, to make a final check andcomparison of the algorithms pro-posed solutions, we used a moredetailed simulation that aims tomore closely approximating the realworld. For example, tests under-taken on public transport timetabledata showed that waiting times canbe significantly decreased in manycases by providing more alternatives

at each stop: in fact our results showed a25% decrease in waiting time in 20% ofthe scenarios considered. Therefore,mainstreaming such policy-based plan-ning approaches in every-day life wouldsignificantly improve the user experi-ence of many PTS.

Link:

http://resweb.watson.ibm.com/researcher/view_group_subpage.php?id=4130

References:

[1] A. Botea, E. Nikolova, M.Berlingerio: “Multi-modal jour-ney planning in the presence ofuncertainty”, in proc. of ICAPS'13,AAAI, 2013.[2] B. Chen, et al.: “Uncertainty inurban mobility: Predicting waitingtimes for shared bicycles and parkinglots” in proc. of ITSC'13, IEEE, 2013.[3] T. Nonner: “Polynomial-timeapproximation schemes for shortestpath with alternatives” in proc. ofESA’12, Springer, 2012.

Please contact:

Tim Nonner, Marco LaumannsIBM Research - Zurich, SwitzerlandE-mail: [email protected],[email protected]

Adi BoteaIBM Research - Dublin, IrelandE-mail: [email protected]

Stochastic travel Planning for Unreliable Public

transportation Systems

by Tim Nonner, Adi Botea, Marco Laumanns

As part of an IBM First-Of-A-Kind project, the IBM research labs in Dublin and Zurich joined forces to

investigate planning approaches for unreliable and highly stochastic public transportation systems (PTS).

Figure 1: An example policy for travelling

from Stop A to Stop D.

http://resweb.watson.ibm.com/researcher/view_group_subpage.php?id=4130

The Smart City concept is on the researchagenda of many European Union (EU)and other international institutions andthink-tanks. As urban populations grow,innovative information and communica-tion technology (ICT) initiatives are seenby many as one of the key factors that willallow modern cities to reach or maintain agood and sustainable quality of life fortheir inhabitants, allowing for the timelyand equitable distribution of resources.

These ICT-based systems are based ondecentralised and distributed designs,comprised of many autonomous and inter-acting entities, known as collective adap-tive systems (CAS). CAS are required toadapt their services seamlessly to thechanging needs of their users, who alsoform an integral part of the system. Theytypically consist of a large number of spa-tially distributed, heterogeneous entitieswith decentralised control and varyingdegrees of complex autonomous behav-iour. This requires the development ofnovel scalable analysis techniques toinvestigate their dynamic behaviour andsupport the design and operational man-agement of a wide range of such systems.In the QUANTICOL project [1], threeprincipal case studies drive the develop-ment of a design and analysis frameworkfor CAS: two smart urban transportationsystems (smart bus systems and bike-sharing schemes) and smart grid applica-tions. We present some of the QUAN-TICOL research performed at ISTI-CNR.

In the first year of the project, we devel-oped several scalable analysis techniquesthat exploit mean field and fluid flow tech-niques, in combination with logic-basedmodel-checking, to support the investiga-tion and prediction of dynamic resourceusage. Mean field techniques were origi-nally developed in the field of statisticalphysics to cope with the analysis of verylarge scale systems composed of inter-

acting objects such as molecules in a gas.The possibly non-linear behaviour of suchsystems is conveniently modelled by adeterministic approximation, i.e., the limitfor an infinite number of agents, given asthe solution of a set of differential equa-tions (in the continuous case) or differ-ence equations (in the discrete case). Theircombined application with model-checking techniques provides a way toverify properties of individual entities inthe context of a large system on whichthey depend, but also properties of theglobal system or combined local andglobal properties. An example is the studyof the potential effects of user-incentiveson maintaining a satisfactory distributionof bikes and empty parking slots overtime. The extension of these techniques toaddress spatial aspects, including spatialmodel-checking, is a major objective ofthe project [2].

A further objective of QUANTICOL is tostudy the relationships between (represen-tations of) small populations and a com-pact (family) representation of a largepopulation ‘built’ from these smaller pop-ulations, by indicating the commonalitiesand variabilities of single entities in theiroverall environment. As an initial step inthis direction, we performed variabilityanalyses on a bike-sharing product line,considering its behaviour to exhibit vari-ability, not only in the kind of featuresinvolved but also in the timing and proba-bility characteristics of these features.

In this context, ISTI-CNR initiated a col-laboration with “PisaMo S.p.A. aziendaper la mobilità pisana”, an in-housepublic mobility company in theMunicipality of Pisa, that had recentlyintroduced a public bike-sharing system(CicloPi) in Pisa. This led to an initial fea-ture model of a family of bike-sharingsystems, annotated with attributes andglobal quantitative constraints aiming to

minimize the total cost of a chosen con-figuration while simultaneously aimingto maximize customer satisfaction andcapacity (of docking stations).

We have studied the specification andanalysis of the possible behaviour of afamily of bike-sharing systems in termsof the capacity of their docking stations ina value-passing modal process algebra,considering a dynamic redistributionscheme as an optional feature. Futurework includes studying a further para-metric extension of the value-passingmodelling and verification environmentas well as the addition of a quantitativedimension to the behavioural model.

QUANTICOL will run until March 2017and is coordinated by Jane Hillston fromthe University of Edinburgh (UK). Otherpartners are EPFL (Switzerland), IMTLucca (Italy), University of Southampton(UK), LMU (Germany), INRIA (France)and ISTI-CNR (Italy). We thank MarcoBertini from PisaMo S.p.A. for gener-ously sharing his knowledge on bike-sharing systems with us.

Links:

QUANTICOL: http://www.quanticol.eu/PisaMo: http://www.pisamo.it

References:

[1] L. Bortolussi et al.: “A QuantitativeApproach to the Design and Analysis ofCollective Adaptive Systems, FoCAS’13,Taormina, Sicily, Italy, 2013.[2] J. Hillston: “Challenges forQuantitative Analysis of CollectiveAdaptive Systems”, TGC 2013, BuenosAires, 2013, Springer LNCS, Vol. 8358,pp.14-21, 2014, DOI: 10.1007/978-3-319-05119-2_1.

Please contact:

Mieke Massink, ISTI-CNR, ItalyE-mail: [email protected]



A Quantitative Approach to the

Design and Analysis of Collective

Adaptive Systems for Smart Cities

by Maurice ter Beek, Luca Bortolussi, Vincenzo Ciancia, Stefania Gnesi, JaneHillston, Diego Latella and Mieke Massink

It’s smart to be fair. Researchers from the Formal Methods and Tools group of ISTI-

CNR are working on scalable analysis techniques to support smart applications for

the efficient and equitable sharing of resources in the cities of our future. The

research is being carried out under the European FET-Proactive project, QUANTICOL.

Figure 1: public bike-sharing system

(CicloPi) in Pisa


We witness a revolution that capitalizeson the prevalence of networked(embedded) devices ranging fromsimple ones, such as sensors, to complexCyber-Physical Systems (CPS), in orderto empower a new generation of innova-

tive anytime-anywhere services. TheCPS revolution enables system inter-connection and fine-grained informa-tion acquisition at unprecedentedscales. The analytics of CPS data pavethe path towards informed real-time

decision making in complex infrastruc-tures. For Smart Grid Cities this impliesinnovative solutions [1] in energy man-agement, sophisticated demand-response (DR) and demand-side man-agement (DSM), optimal resourceusage, reliability and security, new busi-ness opportunities etc. The Smart Gridand its services [2] are seen as an inte-gral part of the Smart City of the future.Several ongoing efforts [1] strivetowards capitalizing on the hyper-con-nected information infrastructure andthe collaboration among the things,services, and systems expected to existin the Smart City.

The SmartKYE project project goesbeyond existing efforts towards a real-istic approach for future Smart GridCities. Most research projects [1] thatdeal with aspects of energy manage-ment are still focusing on core infra-structure issues that will enable them tomeasure energy production and con-sumption and potentially manage it [3].Some go a step further towards buildinganalytics and value-added servicesbased on the data acquired. Recentadvances in cloud computing and highperformance in-memory databasestuned for delivering timely analyticsand various web tools/apps [3]empower such efforts.

One key point that may be problematic,however, is information ownership andmanagement. Most of the approachesdeveloped today assume usually cen-tralized unconditional access to the data[2]. This is a valid assumption withinthe scope of an organization, such as anenergy provider, that controls the wholevalue chain. However, in large-scaleinfrastructures multiple such stake-holders will emerge with portions of thedata, and many are likely to be reluctantto share their data unconditionally sinceit represents a key asset and competitiveadvantage of an enterprise. Hence, busi-ness models need to be extended toinclude interactions among stake-holders that will deliver value-added

Query-Driven Smart Grid City Management

by Stamatis Karnouskos

Cyber-physical systems underpin modern Smart Grids and Smart Cities. New approaches are

required to enable efficient, secure and decentralized use of the huge amount of generated data in

value-added applications. The SmartKYE project (www.smartkye.eu) is investigating such directions

for query-driven integration of Smart Grid City Systems.

Figure 1: The SmartKYE concept of query-driven integration and management

Figure 2: The SmartKYE architecture

services without revealing all of thedata, i.e., that will reveal relevantdetails without handing over an enter-prise’s core advantages.

The SmartKYE project capitalizesexactly on that vision. Contrary to cur-rently overwhelmingly centralizedinformation gathering approaches,SmartKYE considers a query-driveninteraction with the vast number ofstakeholders and their systems (as illus-trated in Figure 1). A loosely coupledinfrastructure enables stakeholders tojoin or leave the SmartKYE system,and the core infrastructure data maystill be owned by the respective sys-tems. To what extent and to what reso-lution such data is communicated toother stakeholders, depends on thequeries issued and the willingness orcontract-based negotiated actions. Bynot owning the data, and having aninfrastructure that intelligently handlesthe communication between consumersand producers of information, scala-bility can be achieved, as well ascoevolution at both ends. As depictedin Figure 1, several stakeholders mon-itor and control parts of the smart gridcity infrastructure. As Figure 2 illus-trates, the acquired information is gath-ered via various Energy ManagementSystems (EMS) which also host controlcapabilities for actuating the under-lying infrastructure.

Each EMS has its own local view andcan operate autonomously from therest. In the cloud, the Open EnergyServices Platform (OESP) acts as theglue among the disparate systems andthe information consumers who, in ourexample case, are portrayed at the topas the municipal information systems.The OESP hosts several advancedfunctionalities including: the capabilityof accepting queries from the end-userapplications, disaggregating them tosee which EMS need to be addressed,contacting the EMS to acquire the nec-essary information, aggregating theanswers, and delivering the informationas requested. The latter is done withadditional indicators on the quality ofdata delivered. In addition, OESPoffers services via which managementactions can be undertaken on the infra-structure. Finally the end-users interactvia cockpits with the system. InSmartKYE, as a proof of concept, twocockpits are addressed with differentcontexts: • the Business Cockpit (BC) targets the

municipal administrators and deci-sion makers who can get high-levelinformation and can be assisted intheir decision processes; and

• the Monitoring and Control Cockpit(MCC) which targets the municipalitytechnical personnel and infrastructuretechnical managers, by providingthem with fine-grained technical

information as well as control (man-agement) capabilities.

References:

[1] V. Giordano, A. Meletiou, C. F.Covrig, et al: “Smart Grid projects inEurope: Lessons learned and currentdevelopments 2012 update,” JointResearch Center of the EuropeanCommission, JRC79219, 2013[2] S. Karnouskos, D. llic, P.Goncalves Da Silva: “Assessment ofan Enterprise Energy Service Platformin a Smart Grid City Pilot”, in IEEE11th International Conference onIndustrial Informatics (INDIN),Bochum, Germany, 2013[3] S. Karnouskos, P. Goncalves DaSilva, D. Ilic: “Developing a WebApplication for Monitoring andManagement of Smart GridNeighborhoods”, in IEEE 11thInternational Conference on IndustrialInformatics (INDIN), Bochum,Germany, 2013.

Link:

SmartKYE Project Web Site:http://www.smartkye.eu

Please contact:

Stamatis KarnouskosSAP, GermanyE-mail: [email protected]



Air quality continues to be a seriousissue for public health, the environmentand ultimately, the economy ofEuropean countries. Poor air qualityresults in ill health and premature deathsand damages ecosystems, crops andbuildings. Urban areas, where themajority of European’s live, are mostseriously affected. In recent years,Europe has significantly reduced theemissions of several air pollutants suchas sulphur dioxide (SO2), carbon

monoxide (CO), benzene (C6H6) andlead (Pb). However, particulate matter(PM), ozone (O3), nitrogen dioxide(NO2) and some organic compoundsstill represent a serious threat. Thereport, “Air Quality in Europe”, pub-lished in October 2013 by the EuropeanEnvironment Agency (EEA) [1]describes the effects of air pollution onhealth, ecosystems and the climate. Italso provides an overview of policiesand measures introduced in Europe to

improve air quality and minimize theimpacts of air pollution impacts.

Real-time, accurate monitoring of airpollution levels in urban areas plays akey role in enabling appropriate andtimely public health decisions to bemade and thus, is of paramount impor-tance. Currently, air quality is typicallymonitored through large and expensivesensing stations, installed at strategiclocations (few) such as intersections.

‘U-Sense’, A Cooperative Sensing System

for Monitoring Air Quality in Urban Areas

by Giuseppe Anastasi, Paolo Bruschi and Francesco Marcelloni

Air quality has a serious impact on public health, the environment and, ultimately the economy of

European countries. In this article we present U-Sense, a cooperative sensing system for real-time

and fine-grained air quality monitoring in urban areas. U-Sense allows for monitoring to occur in

places where people spend the majority of their day-to-day lives.


These stations do allow for accuratemonitoring but their limited spatial cov-erage means this information isrestricted to specific areas. In addition,the sensing stations are managed bypublic authorities which means that thepollution data they gather are often notpublicly available. Citizens, however,are typically interested in knowing theair quality conditions in places whichare relevant to their daily lives such asat home, school, work, or publicspecies.

With this need in mind, we have devel-oped U-Sense, a cooperative sensingsystem which allows for real-time, fine-grained air quality monitoring in urbanareas (Figure 1). U-Sense relies on low-cost sensor nodes, equipped with appro-priate gas sensors, which can be pri-vately installed by citizens. The sensornodes are powered by batteries whichallow for flexible deployment and easyrelocation. Users can share their meas-urements using social networkingwhich enables cooperating sensing.

Data transfer can occur in a number ofdifferent ways. For example, sensornodes can be connected directly to thedatabase through a home-based WiFirouter. Alternatively, data can be firsttransferred to an intermediate relaynode and then on to the database. Thisrelay node can be the user’s smartphoneor, more commonly, any mobile agent(e.g., a mobile relay node mounted on abus or taxi). The sensor nodes can onlytransfer their locally acquired data whenthe relay node comes nearby. Hence, anopportunistic communication protocolis used for data transfer. Data accumu-lated on the database can be used forproviding a number of community serv-ices through a variety of different userinterfaces (e.g., the Web, smartphone ortablet). The range of services envisagedinclude access to air quality data meas-ured by individual sensor nodes at anygiven location, the visualization of pol-lution maps and search facilities for lesspolluted paths. The web-based userinterface for U-Sense is shown inFigure 2.

The U-Sense project is still in progress.We have completed the software imple-mentation component and we havedeployed a number of sensor nodes inacross our city (Pisa, Italy). LibeliumWaspmote sensor nodes have been usedfor this testing phase which are

equipped with a sensor board thatallows for the following air qualityparameters to be measured: CO (carbonmonoxide), CO2 (carbon dioxide), NO2(nitrogen dioxide), O3 (ozone), VOC(volatile organic compound), tempera-ture and humidity.

This research activity is being under-taken as part of the “SMARTY” Project,funded by the Regional Government ofTuscany using European funds. Theproject aims to develop innovative serv-ices for sustainable transport andmobility in smart cities. Along with theUniversity of Pisa, the University ofFlorence and a number of small andmedium enterprises are contributing tothis project.

Reference:

[1] European Environment Agency,“Air quality in Europe - 2013 Report”,October 2013. Available at:http://www.eea.europa.eu/publications/air-quality-in-europe-2013.

Link:

http://www.iet.unipi.it/~anastasi/

Please contact:

Giuseppe AnastasiDept. of Information Engineering,University of Pisa, Tel: +39 050 2217 559E-mail: [email protected]

Figure 2: Web-based user interface for the U-Sense system.

Figure 1: The architecture of the

U-Sense system

http://www.eea.europa.eu/publications/air-quality-in-europe-2013

This research was motivated by thepotential opportunities to use renewable,solar energy in public lighting servicesvia the appropriate combination of LEDluminaries, energy generation andstorage, as well as sensor technologieswith novel data processing, communica-tion and control methods. To this end,an industry-academy consortium formedby GE Hungary, the BudapestUniversity of Technology andEconomics, and two institutes of theHungarian Academy of Sciences (MFAand SZTAKI) developed a so-calledenergy-positive community microgrid(E+grid). E+grid balances energydemand against production and guaran-tees the required level of street lightingeven at times of moderate-durationpower outages. In this article, we outlinethe key aspects of the information, com-munication and control features of theE+grid system.

System architectureThe E+grid system reduces the energyconsumption of street lighting by usingLED luminaries that regulate theirlighting levels according to the sur-rounding environmental conditions, thusproviding just the required level oflighting at all times. This is achieved bymounting motion sensors and smart con-trollers to each light pole, which are thenconnected by wireless communication.A positive yearly energy balance isachieved by photovoltaic (PV) energygeneration, whereas protection againstpower outages is guaranteed by batterystorage. The system has a bi-directionalgrid connection, which enables tradingwith electricity and can be used toexploit variable energy tariffs. Energyflows are monitored by smart meters. Alocal weather station collects weatherdata and hosts a twilight switch whichallows for the lighting periods torespond to the current environmentalconditions. The overall system (Figure1) is monitored and controlled by a cen-

tral computer (CC): black connectorsindicate power flow while red linesshow information flow (smart metersare not depicted).

The central computerThe CC of E+grid, an innovative cloud-deployed software application, is one ofthe key orchestrators of the system,enabling the adaptive lighting behav-iour. Its core functions include:• which are delivered by a web applica-

tion and the extensive use ofJavaScript. The graphical user inter-face (GUI) (Figure 2) providesfriendly access to, among others, thegeographic information system (GIS)of the installed luminaries, the man-agement of the energetic componentsand the visual analysis and export ofthe collected data; and

• which enable the outer platform com-ponents to communicate with the CC.These are provided through a dedicat-ed layer which is responsible both forhandling the connections and validat-ing the semantics. Each component isassisted by a proper data synchro-

nization process, whereas the controlof the outdoor lighting system is del-egated to a specific lighting sched-uler.

Controlling the energy flowOne of the main roles of the CC is con-trolling the energy flow, not only forminimizing the total energy cost, butalso for ensuring the robustness againstpower outages through the use of bat-tery storage. To achieve this, energyproduction and consumption must beforecasted. This is carried out by fittingtime series models to the available data[1]. An assessment of several differentmodels showed that good predictions ofenergy production can be obtainedusing nonlinear autoregressive exoge-nous (NARX) models. These modelsuse wavelet-type nonlinearities, wherethe exogenous components come from aclear-sky model. Efficient energy con-sumption forecasts can also be achievedusing Box-Jenkins type models, wherethe exogenous inputs come from aver-aged historical data. The controlleritself uses a model-predictive approach



Monitoring and Controlling Energy-positive

Public Lighting: the E+grid System

by Balázs Csanád Csáji, Borbála Háy, András Kovács, Gianfranco Pedone, Tibor Révész, and JózsefVáncza

The concepts of smart cities and self-sustaining renewable energy systems are revolutionizing the

world of public lighting. This paper presents the architecture of a novel, adaptive and energy-positive

outdoor lighting system, as well as the IT solutions that control and monitor the operation of the

whole system.

Figure 1: Architecture of the E+grid system

and works on a rolling-horizon whereeach time step involves solving a linearoptimization problem [1].

Prototype systemThe E+grid system design has been val-idated through simulation experimentswhich confirmed that the system can

achieve a slightly positive yearly energybalance. A physical prototype, con-taining 130 luminaries and 152 m2 ofPV panels is currently being deployed atthe campus of the MFA Institute of theHungarian Academy of Sciences. Thisprototype adopts three different batterytechnologies and four types of PV

panels in order to evaluate their long-term performance under real-life envi-ronmental conditions.

AcknowledgementThis project has been supported bygrants from the National DevelopmentAgency, Hungary, under contract num-bers KTIA KMR 12-1-2012-0031 andNFÜ ED-13-2-2013-0002. B. Cs. Csájiacknowledges the support of the JánosBolyai Research Fellowship No.BO/00683/12/6.

Reference:

[1] B. Cs. Csáji, A. Kovács, J. Váncza:“Prediction and robust control ofenergy flow in renewable energysystems”, in proc. of the 19th IFACWorld Congress, to appear, 2014.

Please contact:

András KovácsFraunhofer PMI, SZTAKI, HungarianAcademy of Sciences, HungaryTel: + 36 1 279 6299E-mail: [email protected]


Figure 2: A graphical user interface (GUI) of the central computer

Demand-Side Management in Smart

Micro-Grids: An Optimization Perspective

by Talbi El-Ghazali

With the smart grid revolution, the energy consumption of houses will play a significant role in the

energy system. Indeed, home users are responsible for a significant portion of the world’s energy

needs but market prices for this energy remain inelastic (i.e., energy demands do not follow energy

prices). Thus, the performance of the whole energy generation and distribution system can be

improved by optimizing the management of household energy use. There are a number of

challenges associated with this goal including cost, the environmental considerations, user comfort

and the presence of multiple decision makers (e.g., the end users and energy operators).

The smart micro-grid is an integratedsystem which supports distributedenergy resources and multiple electricalloads but operates as a single power grid[1]. It is a smaller version of the tradi-tional electrical grid or the smart digital-ized grid which working independentlyof, or interconnected with, a largerexisting power network. Smaller scalegrids can deliver a wide range ofimprovements including greater relia-bility, fewer line losses, better fail recov-eries, increased energy efficiencies,carbon emission reductions, reduced

demands on the transmission infrastruc-ture, cost reductions. Finally, it intro-duces the possibility of using alternativeenergy sources because more localizedsources of power generation can berelied on. The global interest inreducing fuel consumption in favor ofrenewable energies has been a catalystfor the growth in the use of smart micro-grids. They are an ideal way to integraterenewable resources at the communitylevel and allow for customer participa-tion in the electricity enterprise. Theyenable electricity to be locally gener-

ated, distributed and electricity flows toconsumers to be regulated.

A smart building system has its ownmicro-grid and some decentralizedresources (e.g., a wind generator, CHPgenerator, boiler, thermal storage andelectrical storage), to provide the basicelectricity needs (Fig.1). It may alsofeature a grid connection which allowsit to obtain electricity during peak hoursor sell electricity (back to the grid)when surplus electricity is generated.Electric micro-grids are also regarded

as a security mechanism as they ensurehomes have access to energy duringevents such as weather-related black-outs or military attacks. All of thesebenefits are stimulating an increaseddemand worldwide for micro-gridswhich are being applied in a variety ofcontexts including campus and residen-tial environments, military sectors andcommercial and industrial markets. Theglobal market for micro-grids topped$5billion in 2011, according to the 2013Microgrid Market Report and is itseems likely this figure will reach $27billion by 2022.

Optimization perspectiveWith the Smart Grid revolution, theenergy consumption of households willplay a significant role in the energysystem. Indeed, home users are respon-sible for a significant portion of theworld’s energy needs but market pricesfor this energy remain inelastic (i.e.,energy demands do not follow energy

prices). Thus, the performance of thewhole energy generation and distribu-tion system can be improved by opti-mizing the management of householdenergy use. Smart Grid technologieshave been deployed in a variety of con-texts including smart sensors on distri-bution lines, smart meters in homes,advanced metering, transmission effi-ciencies, smart switches, enhanceddemand response capabilities, distribu-tion automation, electric vehicle inte-gration and integrated communicationsystems. These pave the way for thecreation of smart self-managing energygrids within a time varying pricingscheme (i.e., where prices may vary

according to time of day). The SmartHomes of the future will includeautomation systems which may simul-taneously provide lower energy con-sumption costs and a comfortable andsecure living environment for end users.Residential users are expected to play akey role in improving network effi-ciency,through the adoption of intelli-gent mechanisms that manage energydemand. These days, most electricityconsumers act as price takers with flatrates and no acknowledgement of thedifferences that exist in electricityprices. Therefore, they don’t have anyincentives to adjust their electricity con-sumption patterns. However, consumerscan optimally adjust their energy con-sumptions by participating in a housemanagement program which seeks tominimize electricity bills. To achievethis end goal, consumers must respondto shifting electricity prices andredesign their consumption patters tomatch with periods of lower electricity

prices. The user can define the earliestand the latest starting times that aservice can be processed, with a utilityvalue identified for each time slot whichindicates his or her comfort expecta-tions. Thus, the challenge for the house-hold energy management is to calculatewhen the optimal starting point occursacross the different household appli-ances in order to minimize energy costsfor the user whilst maximize their satis-faction. In addition to reducing theenergy bills of single users, joint man-agement of multiple users can also beadvantageous. If energy consumptiontasks occurring across multiple homescan be scheduled, based on the suite of

users’ requirements, there is potentialfor reducing ‘peak demand’ periods. Asall electricity grids require a minimumenergy generation capacity so as to pre-vent failures and power cuts, overallgrid costs might be reduced if peakdemands can also be reduced. Peakshaving, as this phenomena is known,reduces the burden on the central gridand any upgrade expenses which maybe incurred to ensure the grid can fulfillincreasing energy demands. Finally, italso offers savings to both the utilitycompanies and customers by reducingthe need for generation capacity andenergy amounts the utility companiesmust purchase on the open marketduring peak demand periods to preventfailures and power cuts. Thus, reducingthe global contractual peak would bringmultiple advantages, not just to the elec-tric grid but also to residential users andenergy retailers. In the DOLPHINproject, these challenges have been for-mulated as multi-objective optimizationproblems.

Energy prices have to be defined notonly to retrieve their production costsbut also to understand consumerbehavior. Consumers choose their serv-ices or energy providers with the aim ofminimizing their disutility values. Afailure to recognize this may lead to sig-nificant revenue losses. To ensure thishierarchical and multi-objective deci-sion process is captured, where theleader (energy provider) explicitly takesthe reaction of a follower (consumer)into account in their decision process,the energy pricing and managementproblems can be formulated as bi-levelmulti-objective problems [2].

References:

[1] J. Ekanayake et al: “Smart grid:Technology and applications”, Wiley2012.[2] E. Alekseeva, L. Brotcorne, E-G.Talbi: “A new exact method finding afeasible solution for the bi-level energypricing problem with two objectives atthe lower level”, in META’2014,Marrakech Morocco, October 2014.

Please contact:

Talbi El-Ghazali University of Lille, Inria, CNRS,FranceE-mail: [email protected]



Figure 1: Demand-side management in a smart home.


The introduction of new kinds of energymixes to the electricity grid is a chal-lenging environmental task for presentand future generations as they fight thepollution and global warming issuesassociated with urbanization. Individualappliances and whole buildings that con-tinuously incorporate local intelligencewhich originates from the new technolo-gies of Internet Of Things (IoT) are thenew infrastructure that this integrationwill be based on. Smart Electricity Gridsare becoming more intensively inte-grated with tertiary building energymanagement systems and distributedenergy generators such as wind andsolar. Optimal energy managementstrategies require the ability to controland predict energy consumption incor-porating all types of DER (distributedenergy resources), at both local andglobal scales.

The INERTIA projectIn the context of the EU FP7- ICTproject INERTIA, a consortium of 10complementary partners from six dif-ferent European Countries aims ataddressing the important challenges thatfuture energy systems will face bybuilding a new modelling methodologythat integrates physical components andcyber technologies, thus creating CyberPhysical Energy Systems. TheINERTIA project works on the defini-tion of a new data management infra-structure to allow electricity productionand consumption to be measured,reported and controlled, delivering anefficient Distribution Grid overlay con-trol/management infrastructure (Figure1). This new infrastructure will maxi-mize the response capacity of the vast,small-commercial prosumer base (e.g.,tertiary buildings, offices, etc.), pre-senting incentives and delivering bene-fits through their automated active par-ticipation in the energy market.

INERTIA extends the DSM strategies byincorporating various types of DER

going beyond simple consumer loads,and treats both local generation andconsumption under a single unifiedframework. The concept of DER flexi-bility will also be extended to incorpo-rate various local context parameters,which affect DER operation and per-

formance, as well as potential capacityof DER to provide flexible services tothe Distribution Grid. This modelincludes the impact of communicationnetworks and further, cyber componentsas well as the relevant information ofthe physical system.

Cyber Physical Systems give Life

to the Internet of Energy

by Giampaolo Fiorentino and Antonello Corsi

A new overlay network which runs on top of the electricity grid, puts together old and new renewable

energy sources, giving life to the Internet of Energy. A coordinated and controlled Internet of Things will

optimize the Distribution Grid Control and the Demand Side Management (DSM) based on business,

comfort and occupancy dynamic, introducing the autonomous intelligent Commercial Prosumer Hub.

Figure 1 - The concept behind the INERTIA project. The overlay network will have the task of

balancing distributed energy resources (micro-generation, RES, energy storage systems and

demand) in the grid through the intelligent mining of energy flexibility, as a balancing power

between local generation and local demand.

Figure 2: Holistic flexibility engine - This component is able to integrate and provide forecast

energy consumption providing a real time tool for integrating medium voltage points of the net

in the new smart grid energy.

The INERTIA frameworkThe old approach [1, 2] was to try toalign consumption by asking consumersto reduce their power usage rather thanincreasing the power generation facili-ties. Under this approach, consumersthat cooperated might receive incentivepayments from the power company. Incomparison, the INERTIA projectclaims to solve this issue using IoT tech-nologies to expand DR services basedon the analysis of users’ historical inter-actions with the lighting, HVAC andother appliances controls. This has beenrealized calculating energy consump-tion, flexibility and forecast demandusing a holistic flexibility model, whichcalculates several key performance indi-cators which can be categorized inBusiness, Energy, Comfort andFlexibility. Furthermore, real-timeemployee activities including arrivaland departure times and meetings atten-dance (Figure 2) are also simulated.

Besides, INERTIA takes into accounthow demand response can change theend user comfort using a Bayesian for-malism. This formalism is applied toinfer the probability that any environ-mental situation should be consideredby the user as environmentally uncom-fortable, defining the upper and lower

limits of the energy flexibility that canbe offered to the whole energy system.

So, the holistic flexibility model is ableto safeguard the electricity grid, facing atrade-off between energy provided bythe local generation facility and con-sumers’ consumption. In this respect,the holistic flexibility model generatesthe suitable energy demand strategydeploying DSM operations, which takesinto account the final users comfort.The comfort is evaluated also by thefacility manager selected to enroll aspecific DR program

ConclusionNew IoT technologies are essential forimplementing electrical power systemsin Smart Cities and reducing environ-mental impacts and social costs. At themoment, appliances can self-adjust toconsume less power or even turn them-selves off. Using the data collected fromall the devices, algorithms can calculatethe exact energy demand so to reduce theneed for standby generation. Over time,generation can be automatically shiftedaccording to predicted increases anddecreases in demand. Morover, a frame-work that receives input from relevantsources such as weather information,heating, cooling, lighting equipment and

usage scenarios (occupancy rates) can beimplemented to compute energy con-sumption and identify available flexibili-ties from DER. This could be conductedin a non-invasive way that would respectthe comfort of the end user.

AcknowledgmentsThis work was partially supported bythe EU project “INERTIA - IntegratingActive, Flexible and ResponsiveTertiary Prosumers into a SmartDistribution Grid” (FP7-ICT-2011-8,Grant Agreement No. 318216).

Link: http://www.inertia-project.eu/

References:

[1] A Di Giorgio, L Pimpinella: “An eventdriven Smart Home Controller enablingconsumer economic saving andautomated Demand Side Management” [2] P Palensky, D Dietrich: - IndustrialInformatics, IEEE …, 2011 - Demand sidemanagement: Demand response,intelligent energy systems, and smart loads

Please contact:

Giampaolo Fiorentino, Antonello CorsiR&D Lab - Engineering IngegneriaInformatica S.p.A, ItalyE-mail: {giampaolo.fiorentino,antonello.corsi}@eng.it



The possibilities enabled by Informationand Communication Technologies(ICTs) are driving the evolution andtransition of cities to Smart Cities. Theultimate goal is to increase the aware-ness of citizens’, companies’ and author-ities’ and improve their quality of lifewhile also making it sustainable. A con-siderable number of research directionsembrace Smart Cities: users' privacyprotection [1], detection of maliciousactions and misuses and users' aware-ness through social media. Moreresearch efforts are dedicated to specificfeatures of a Smart City. As an example,the energy forecast techniques used to

predict consumption and allow theusage of alternative energy resources(e.g., solar or wind power) to be sched-uled. What all these research fields havein common is their dependency on the(possibly sensitive) data produced bythe devices forming the Internet ofThings (IoT) of a city. The possibilitiesenabled by Smart Cities demand fornovel data processing paradigms toform the expertise of public and privatecompanies. Based on our experiencewith both academic and industrial part-ners, in this article we discuss some ofthe challenges associated with data pro-cessing in Smart Cities.

Scalable and online access to thedata In a Smart City, millions of messageswill be exchanged on a daily basis byhundreds of thousands of devices (e.g.,mobile phones, electrical meters,weather stations, etc.). For example,more than 1.2 million messages areexchanged on a daily basis within anAMI infrastructure (owned by one ofour industrial partners) that covers ametropolitan area with roughly 600,000inhabitants [2]. The information gener-ated by such devices could be matchedand joined to enhance the managementof Smart Cities. For example, energy or

When Smart Cities meet Big Data

by Vincenzo Gulisano, Magnus Almgren and Marina Papatriantafilou

Sharing information is a key enabler in the transition of a city becoming smart. Information,

generated by the ICT backbone of a city, and maintained by distinct public and private entities,

comes with processing challenges that must be addressed in order to increase citizens’ quality of

life and make their cities sustainable. In CRISALIS and SysSec, we investigate such challenges

from a security perspective in order to protect and enhance smart cities’ sensitive infrastructures.


water losses caused by faulty devices

could be reduced by matching the con-

sumption measured by users’ meters

with the one measured by other utilities’

systems. To this end, on-the-fly pro-

cessing of data becomes all the more

important while traditional store-then-

process approaches in which each com-

pany retrieves its data and stores it in

order to access it sometime in the future

might be no longer appropriate.

Think in a distributed and parallel

fashion

Smart Cities will be composed of sev-

eral independent networks (even within

the same stakeholder). Hence, no cen-

tralized application will embrace the

information carried by the messages

exchanged by the devices. At the same

time, the huge volume of information

shared by ICT devices will make par-

allel processing a necessity [3]. To this

end, pushing the analysis closer to the

sources of information would be a nat-

ural way of analyzing the messages

exchanged by them and leverage the

information they carry. Challenging

aspects in this context will be imposed

by the constrained resources of such

devices.

Validate, estimate and protect the

data

Cheap, resource-constrained devices

are largely employed to build the net-

works that will form the IoT of a Smart

City. Unfortunately, the data measured

and reported by such devices (e.g.,

energy consumption readings) is usu-

ally noisy and lossy. Reasons of this are

not limited uniquely to the devices

themselves (e.g., faulty or badly cali-

brated devices, lossy or overloaded

communication channels) but can also

be caused by (possibly malicious) citi-

zens. As an example, incorrect con-

sumption readings could be manipu-

lated by malicious users aiming to

lower their bills. To this end, validation

schemes, estimation schemes and secu-

rity countermeasures must be adopted

in order to ensure that who leverages the

information is not mislead by incorrect,

partial or malicious data.

The shift from cities to Smart Cities

depends on the efficiency with which

information is shared among citizens

and private and public companies. This

information brings challenges, and, fol-

lowing the big data revolution, novel

processing schemes must be adopted to

enable the possibilities that exist of this

domain. All the possibilities enabled by

smart cities, like improved quality of

life or energy efficiency, shall build on

top of efficient data processing and

users’ privacy protection schemes.

CRISALIS may be contacted at con-

[email protected]. SysSec may

be contacted at the corresponding con-

[email protected], followed in

twitter (twitter:syssecproject) and

Facebook

(http://www.facebook.com/SysSec).

References:

[1] V. Tudor, M. Almgre, M.

Papatriantafilou: “Analysis of the

Impact of Data Granularity on Privacy

for the Smart Grid”, in proc. of

WPES’13

[2] Z. Fu et al.: “Managing your Trees:

Insights from a Metropolitan-Scale

Low-Power Wireless Network”, in

proc. of CCSES’14, held in

conjunction with the INFOCOM 2014

[3] V. Gulisano, M. Almgren, M.

Papatriantafilou: “METIS: a Two-Tier

Intrusion Detection System for

Advanced Metering Infrastructures”, in

ACM e-Energy, 2014.

Please contact:

Vincenzo Gulisano

Chalmers, Sweden

E-mail: [email protected]

Figure 1:

Key enablers

cItyAM: Managing Big Urban Data

for Analyzing and Modelling Cities

by Alessandro Bozzon, Claudia Hauff and Geert-Jan Houben

As we race towards fully connected living environments at full speed, urban stakeholders and decision

makers are demanding analytic solutions that are able to provide actionable insights about citizens

and their interactions with these environments. The Web Information Systems group at Delft

University ofTechnology responded to this challenge by developing cItyAM, an integrated platform that

supports the analysis and modelling of urban data. This data is received from a wide range of sources

including social media, (social) sensors, open government data and a multiple knowledge repositories.

An open and extendable urban analytics

platform, cItyAM is designed to cope

with big urban data, providing stake-

holder and decision makers with a set of

tools to 1) integrate and analyze various

urban data types and 2) model, engage

and interact with people in urban spaces.

In view of the trend toward making

cities fully connected and rising urban

population densities, cItyAM provide

the wherewithal to exploit big urban

data, thus allowing urban phenomena

to be mapped (e.g., the mobility, envi-

ronmental, sustainability and social

aspects associated with urbanlifestyles).

The cItyAM platform has been devel-oped as part of a research initiativestarted by the Web Information Systemsgroup at the Delft University ofTechnology in 2012. It was driven and

inspired by several European- andnational-level initiatives including theCity Data Fusion EIT ICT Lab project(conducted in collaboration withPolitecnico di Milano, CNR Pisa,SIEMENS, and Telecom Italia), theAmsterdam Institute for AdvancedMetropolitan Solutions (AMS) initia-

tive (which includes WAGENINGENUR, MIT, TNO, and other societal andindustrial partners), the SHINE project,the ImReal EU FP7 project, and theCOMMIT SEALINCMedia project(conducted in collaboration with the VUAmsterdam, CWI, the Rijksmuseum,and others), as well as the experiencegained from the spinoff activities aroundTwitcident. The goal of the cItyAM plat-form is to progress urban data manage-ment, by advancing our ability to “feelthe pulse” of cities: it provides method-ologies and systems for data fusion,analysis and visualization.

The cItyAM platform capitalizes on theexperience of the Web InformationSystem group and employs four mainpillars:1. Semantic user modelling techniques,

drawn from the experiences gainedduring the ImReal project [1], and inthis project tailored to the needs andfeatures of urban citizen modelling. Itallows for the semantically enrichedanalysis of conversations and activi-ties which take place on social mediachannels (e.g., Twitter, Instagram andFoursquare).

2. Crowdsourcing and crowdsensingtechniques for data collection, refine-ment and verification, integration andinterpretation and processing. City-associated data can be patchy, noisyand disparate. Creating city relatedknowledge is driven by an extendableset of social sensing tools, like theCrowKnow system. Automatic tech-niques for data cleansing and linkingoften fail to achieve the levels ofaccuracy required for the effectivefusion of urban data. Crowd-sourcingthis cleansing and linking function isemerging as an effective means tosolve the task when automatic meth-ods fail.

3. Expert finding and crowd engage-ment strategies, developed within theCOMMIT SEALINCMedia project[2] which actively drive the routingand engagement of citizens uponrequest from stakeholders. An exam-ple of this is when a crowdsensingcampaign is initiated. The cItyAMincorporates incentive and engage-ment techniques that are specificallydesigned for urban environments andlife styles.

4. Using the semantic data integrationtechniques that were also developedin the context of the ImReal project,



Figure 2: The Amsterdam Social Dashboard, developed on top of the cItyAM platform for the

Amsterdam Institute for Advanced Metropolitan Solutions (AMS) opening day.

Figure 1: The role of the cItyAM platform


cItyAM can interface with a broadrange of data sources. The RDFGearsframework [3] allows for the integra-tion, publication and retrieval of datacoming from physical sensors, crowdsand existing knowledge repositoriessuch as DBpedia and Freebase.

Selected components of the cItyAMplatform have been deployed in the lasttwo years to monitor and analyze city-scale events like the Milano DesignWeek, the Lucca Comics & Games, theDutch King’s Day and others. In thefuture, we plan to deploy cItyAM as acontinuous service operating in multiplecities around the world. In the context ofAmsterdam, for instance, we plan tointegrate with urban data sources man-aged by the municipality to supportstudies on urban mobility and social city

life. In Rotterdam, the plan is to createsocial sensing solutions for water man-agement, with a specific focus on emer-gency management for flooding events.Finally, we plan to provide a 24/7 socialrecommendation service covering theEXPO 2015 event in Milan.

Links:

http://citydatafusion.orghttp://www.ams-amsterdam.com/http://direct.tudelft.nl/shine-117.htmlhttp://www.imreal-project.euhttp://www.commit-nl.nlhttp://twitcident.comhttp://www.wis.ewi.tudelft.nl/CroKnowhttp://www.wis.ewi.tudelft.nl/research-lines/rdfgears

References:

[1] Qi Gao, et al.: “A ComparativeStudy of Users’ MicrobloggingBehavior on Sina Weibo and Twitter”,UMAP 2012[2] J. Oosterman, et al.: “Crowd vs.experts: nichesourcing for knowledgeintensive tasks in cultural heritage”,WWW (Companion Volume) 2014[3] M. Grabowski, J. Hidders, J. Sroka:“Representing MapReduce,Optimisations in the Nested RelationalCalculus”, BNCOD 2013

Please contact:

Alessandro Bozzon, Claudia Hauff,Geert-Jan HoubenTU Delft, The Netherlands E-mail: [email protected],[email protected],[email protected]

Urban-Scale Quantitative Visual Analysis

by Josef Sivic and Alexei A. Efros

Urban-scale quantitative visual analysis opens up new ways Smart Cities can be visualized,

modelled, planned and simulated by taking into account large-scale dynamic visual inputs from

a range of visual sensors.

Map-based street-level imagery, such asGoogle Maps with Street View, providesa comprehensive visual record of manycities around the world. For example,the visual appearance of Paris has beencaptured in almost 100,000 publicallyavailable Street View images. We esti-mate there are approximately 60 millionStreet View images for France alone,covering all the major cities. In the nearfuture, additional visual sensors arelikely to become more wide-spread, forexample, cameras are being built intomost newly manufactured cars. Anotherincreasing trend is the ability for indi-viduals to continuously capture theirdaily visual experiences using wearablemobile devices such as the GoogleGlass. Collectively, all this data provideslarge-scale, comprehensive and dynami-cally updated visual records of urbanenvironments.

Automatic analysis of urban visual dataThe next opportunity lies in developingautomatic tools that allow for the large-scale quantitative analysis of the avail-able urban visual records. Imagine a sce-nario where we could provide quantita-tive answers to questions like:

• What are the typical architectural ele-ments that characterize the visualstyle of a city (e.g., window or bal-cony type)? (Figure 1a)

• What is their geo-spatial distribution?(Figure 1b)?

• How does the visual style of an areaevolve over time?

• What are the boundaries betweenvisually coherent areas in a city?

These examples just touch on the rangeof interesting questions that can beposed regarding the visual style of acity. Other types of questions concernthe distribution of people and theiractivities For example, how does thenumber of people and their activities ata particular place evolve during a day,the seasons or years? Or perhaps youmight want to know the make-up ofactivities on a given street: the presenceof tourists sightseeing, locals shopping,the elderly walking their dogs, or chil-dren playing. This type of data can alsobe used to respond to significant urbanissues, for example, what are the majorcauses of bicycle accidents?

New applicationsTo progress the way we can respond tothese types of questions would open-upnew ways Smart Cities can be visual-ized, modelled, planned and simulatedby taking large-scale dynamic visualinputs from a range of visual sensorsinto account. Some examples of howthis data might be applied include: • the real-time quantitative mapping

and visualization of existing urbanspaces [1] to support architects anddecision makers (Figure 1),

• the ability to predict and model theevolution of cities [3] (e.g., land-usepolicies and the way they impact onthe visual appearances of differentneighbourhoods),

• obtaining detailed dynamic semanticcity-scale 3D reconstructions andusing them to simulate different envi-ronmental scenarios, e.g., levels ofnoise, energy consumption or illumi-nation, and

• the analysis of human activities, e.g.,evaluating the future success of arestaurant or the need of to introducenew traffic security measures.

The challenge of urban-scale visualanalysisImpressive demonstrations of theanalysis of large-scale data sets havealready started to appear in other scien-tific disciplines. In natural languageprocessing, an analysis of more than 5million books published between 1800and 2000 revealed interesting linguistic,sociological and cultural patterns [2]. Inthe visual domain, however, a similarlarge-scale analysis has yet to bedemonstrated. As visual data and com-putational resources are becoming morewidely available, the key scientificchallenge now lies in developing pow-erful models which can competentlymeet the spatio-temporal, widely dis-tributed and dynamic characteristics ofthis visual data. For example, while thevocabulary and grammar of written textare well defined, there is no acceptedvisual vocabulary and grammar thatcaptures the subtle but important visualdifferences in architectural styles, or thedifferent visual appearances of humanactivities on city streets.

Example: quantitative analysis of architectural styleIn this first phase of investigation, weconsidered quantitative visual analysisof architecture style [1]. Using the largerepository of geo-tagged imagery, wesought to find a way of automaticallyidentifying which visual elements, e.g.,windows, balconies and street signsdefine a given geo-spatial area (in this

case Paris). This is a tremendously diffi-cult task as the differences between thedistinguishing features of differentplaces can be very subtle. We were alsofaced with a difficult search problem:given all the possible patches in all thepossible images, which patches are bothgeographically informative and occurfrequently? To address these issues, weproposed a discriminative clusteringapproach which took into account theweak geographic supervision. We showthat geographically representativeimage elements can be discovered auto-matically from Google Street Viewimagery in a discriminative manner. Weapplied the algorithm on image datasetsfrom 12 cities (Paris, London, Prague,Barcelona, Milan, New York, Boston,Philadelphia, San Francisco, San Paulo,Mexico City and Tokyo), with eachdataset featuring approximately 10,000images. An example of the results wasdiscussed above (and illustrated inFigure 1). This example demonstratesthat these learnt elements are visuallyinterpretable and perceptually geo-informative. We further demonstratethat the identification of these elementscan support a variety of urban-scalequantitative visual analysis tasks, suchas mapping architectural correspon-dences and influences within and acrosscities, or finding representative ele-ments at different geo-spatial scales [1].

Link:

CityLab@Inria Project Lab on SmartCities: http://citylab.inria.fr

References:

[1] C. Doersch, S. Singh, A. Gupta, J.Sivic, and A. Efros. What makes Parislook like Paris? ACM Transactions onGraphics (SIGGRAPH 2012)[2] J.B. Michel et al. Quantitativeanalysis of culture using millions ofdigitized books. Science,331(6014):176–182, 2011[3] C.A. Vanegas et al. Modelling theappearance and behaviour of urbanspaces. Computer Graphics Forum,29(1):25-42, 2010.

Please contact:

Josef SivicInria and Ecole Normale Supérieure,FranceE-mail: [email protected]

Alexei A. EfrosUC Berkeley, USAE-mail: [email protected]



Figure 1: Quantitative visual analysis of urban environments from street view imagery [1].

1a: Examples of architectural visual elements characteristic of Paris, Prague and London, identified through the analysis of thousands of Street

View images. 1b: An example of a geographic pattern (shown as red dots on the map of Paris) of one visual element, balconies with cast-iron

railings, showing their concentration along the main boulevards. This type of automatic quantitative visual analysis has a potentially significant

role in urban planning applications.


Since 2007, the widespread use ofmobile devices (e.g., phones, tablets)and emergence of a new generation ofdevices which are equipped with moreprecise sensors and cameras has placedaugmented and virtual-reality technolo-gies within reach. Such applications arevery resource-demanding and criticallyinterdependent on the quality of user-location information and environmenttracking (e.g., buildings, monuments).Several mobile services based on aug-mented-reality technologies havealready been developed. These can aidnavigation, offer context-aware accessto a wealth of geo-tagged content relatedto cultural heritage, provide informationon city events and assist with urbanplanning issues. However, these servicesstill lack enhanced and robust trackingtechniques that are sufficiently effectiveenough to make the applications reliableand widely deployed.

Today, one of the main challenges inmobile augmented reality applicationsdesign is understanding how our percep-tion of reality can be profitably aug-mented and how this augmentation canbe made to fit seamlessly with the user’sinteraction with the world. TheEuropean VENTURI project whichbegan in 2011 is aiming to develop the

first generation of ubiquitous aug-mented reality (AR) tools that meet realuser needs and fit within the contextthey operate in. To illustrate the poten-tial impacts such technologies mighthave in our everyday lives, a demon-stration will be conducted in front of alarge audience in late 2014. Thisdemonstration will be a cultural-her-itage visit of Grenoble (France) pre-sented on smartphones (Figure 1). Theapplication features an augmented tourof Grenoble’s antique neighbourhood asfar as the 19th century fort (476m eleva-tion). Along the route, the applicationbehaves like an audio AR guide, deliv-ering walking directions and historicalinformation. It also includes a set ofimmersive virtual history galleries.Overall, it exposes the user toGrenoble’s past and present by com-bining 3D-reconstructed monumentsand old geo-positioned pictures. Forexample, the city’s old roman fortifica-tions are overlaid on a live camera viewusing visual-recognition techniques.Finally, to further enhance user immer-sion, the application provides different3D soundscapes synchronized with thevisual rendering.

This demonstration addresses a widespectrum of the technological chal-

lenges a project faces including audio-visual scene analysis to understand theuser’s context, the collection, creation,fusion and delivery of AR content, 3Daudio rendering, mobile human-machine interactions and finally theprovision of a high-accuracy localiza-tion system. To this end, the VENTURIproject consortium brings together dif-ferent ERCIM members such as Inria,Fraunhofer, FBK, mobile device manu-facturers SONY and ST Microeletronicand software companies like MetaioGmbH and EDIAM Sistemas. Thecommon goal of all these partners is todesign a hardware and software plat-form dedicated to such applications.

Of these technologies, the high-accu-racy localization is central to allowingfor smooth integration with vision-based techniques. The TyReX researchteam, located in the Inria Grenoble -Rhône-Alpes Research Centre, Francehas been working since 2010 on thedesign of such a localization system,commonly known as Pedestrian Dead-Reckoning (PDR). This system is basedon the traditional micro-electro-mechanical systems (MEMS) used insmartphones. The on-board accelerom-eter and compass are used to provide anestimation of the user’s relative position

Mobile Augmented Reality Applications

for Smart Cities

by Mathieu Razafimahazo, Nabil Layaïda, Pierre Genevès and Thibaud Michel

The TyReX research team at the Inria Grenoble - Rhône-Alpes Research Centre, France is working on

the design of a high-accuracy localization system. This key technology allows for a smooth

integration of vision-based techniques in augmented reality applications [1].

Figure 1: Examples of

user interfaces seen as

part of the cultural-

heritage visitor

application developed for

Grenoble which uses

augmented reality tools.

and orientation in ‘urban canyon’ envi-ronments (defined as areas with poorGPS-signal strength such as indoors).Combined with geographical data, thiscomponent can also be used to identifyuser-activity patterns (e.g., walking, run-ning or being in an elevator). The inter-pretation of sensor values, coupled withdifferent walking models, allows us toensure continuity in determing the user’slocation when they are both indoors andoutdoors. However, dead recknoning issubject to cumulative errors that aredriven by multiple factors (e.g., sensordrift, missed steps or poor estimation ofstride length). These errors can bereduced by fusing various external sig-

nals from sources such as GPS and Wi-Fi or by relying on the analysis of auser’s trajectories with the help of astructured map of the environment.

PDR has played a key role in the VEN-TURI application since the delivery ofits AR content is mainly based on theuser's estimated location. PDR and theother technologies mentioned in thisarticle aim to provide available infor-mation in a ‘user-centric’ way asopposed to a ‘device-centric’. The 2014demonstration in Grenoble, which com-bines advanced AR, indoor and outdoorPDR and 3D vision tracking will be thisproject ‘s final deliverable.

Links:

TYREX team: http://tyrex.inria.frVENTURI project: http://venturi.fbk.eu

Reference:

[1] J. Lemordant, T. Michel, M.Razafinahazo: “Mixed RealityBrowsers and Pedestrian Navigation inAugmented Cities” in proc. of theGraphical Web Conference, Oct 2013,San Francisco, USA,http://hal.inria.fr/hal-00872721/en

Please contact:

Nabil LayaïdaInria Rhône-Alpes, FranceE-mail: [email protected]



Monitoring People’s Behaviour using Video

Analysis and trajectory Clustering

by Francois Bremond, Vania Bogorny, Luis Patino, Serhan Cosar, Guido Pusiol and Giuseppe Donatiello

Activity discovery within transportation systems (for example subways and roads) or home care monitoring

based on cognitive vision and data-mining technologies, are the core activities of a project at Inria.

It is well known that video cameras pro-vide one of the richest and most prom-ising sources of information aboutpeople’s movements. New technologieswhich combine video understanding anddata-mining can analyse people’s behav-iour in an efficient way by extractingtheir trajectories and identifying themain movement flows within a sceneequipped with video cameras. Forinstance, we are designing an activityrecognition framework which can mon-itor people’s behaviour in an unsuper-vised manner. For each observed person,the framework extracts a set of space-time trajectory features describinghis/her global position within the moni-tored scene and the motion of his/herbody parts. Based on trajectory clus-tering, human information is gathered ina new feature that we call PerceptualFeature Chunks (PFC). The set of PFC isused to automatically learn the particularregions of a given scene where impor-tant activities occur. We call this set oflearned scene regions the topology.Based on a k-means algorithm, a clus-tering procedure over the PFCs is per-formed in order to construct threetopology layers, organized from coarsestto finest. Using topologies and PFCs, weare able to break the video into a set ofsmall events or primitive events (PE),each of which has a semantic meaning.

The sequences of PE and the threelayers of topology, are used to constructa hierarchical model with three granu-larity levels of activity.

The proposed approach has been exper-imented in collaboration with NiceHospital within the FP7 Europeanproject DEM@CARE collectingdatasets to supervise patients suffering

from dementia. These datasets containolder adults performing everyday activ-ities in a hospital room, equipped with amonocular and RGBD cameras (resolu-tion: 640x480 pixels). The activitiesconsidered include ”watching TV“,”preparing tea“, ”answering phone“,”reading newspaper/magazine“,”watering plant“, ”organizing the pre-scribed drugs“, ”writing a check at thedesk“ and ”checking bus routes on a

map“. The mono-camera dataset con-sists of 41 videos and the RGBD datasetcontains 27 videos. For each person, thevideo lasts approximately 15 minutes.For the monocular camera, persondetection is performed using an exten-sion of the Gaussian Mixture Modelalgorithm for background subtraction.For the RGBD camera, we used aperson detection algorithm that detects

people’s heads and shoulders.Trajectories of people in the scene areobtained using a multi-feature algo-rithm that uses features such as 2D-size,3D-displacement, colour histogram,dominant colour and covariancedescriptors.

Experimentations on both datasetsshow that the framework achieves ahigh rate of True Positives and a low

Figure 1: Extracted trajectories of a person at home during the preparation and eating of a

meal (left )and people in a metro station while buying tickets (right).


rate of False Negatives. In total, 99% ofthe performed activities (in real-life) arerecognized by the framework [1].Furthermore, the duration of the recog-nized activities is matched with morethan 80% accuracy to the groundtruthed activities, which means that notonly can the system count the amount ofactivity instances but also pretty accu-rately detect their duration’s. Althougha few activities were missed because offailure to detect finer motions, theexperimental results show that thisframework is a successful system thatcan be used to automatically discover,learn and recognize Activities of DailyLivings (ADLs). In addition, it can beobserved that the framework is useful inmedical applications for supporting theearly diagnosis of Alzheimer ordementia in older adults. The frame-work can successfully distinguishpeople suffering from Alzheimer fromthose with Mild Cognitive Disorder orNormal Controls.

This framework can also be used inmany other fields, such as the video sur-veillance of metros and roads. We haveapplied a variant of this framework intwo scenarios within the FP7 Europeanproject VANAHEIM. The first scenariorelates to the monitoring of activities indifferent locations (e.g., the entrance

concourse) of Paris and Turin metro sta-tions. The results obtained show whichzones have the most intense activities(called Heat Map). In the scene underobservation, rare or uncommon behav-iours such as “jumping above the bar-rier”, “fainting” and “loitering” or fre-quent behaviours such as “buyingtickets” were identified. In the secondscenario, a road lane reserved for buseswas surveyed. Again, we were able tolearn the topology of the scene andreveal which were normal activities(i.e., the passage of a bus into the zone)and abnormal activities (i.e., the pas-sage of other vehicles into the reservedlane). When the results were ground-truthed, a high level of recall and anacceptable degree of precision wereobtained [2, 3].

A variety of domains can benefit fromsmart video analysis. For instance, inthe near future, more than 2 billionpeople will be over 65 years old, andvideo analysis has the potential to helpaging adults in their daily life throughthe use of smart home environmentsand to help providing doctors withactivity observations that can be used todetect possible anomalies for diseaseprevention. Movement analysis incrowded areas, such as metro stations,can detect and alert anomalous behav-

iours, identifying simple issues such asa broken ticket machine to more com-plex events like a robbery or terrorismact. In supermarkets, the analysis ofcustomer movements can provideinformation on how to enhance theshopping experience. However, thisanalysis still has many remaining chal-lenges, such as the accuracy of activityrecognition within huge amounts ofnoisy metadata.

References:

[1]. G. Pusiol, F. Bremond and M.Thonnat: “Unsupervised Discovery,Modeling and Analysis of long termActivities, in proc. of ICVS 2011,Sophia Antipolis, France, 2011. [2] L. Patino, F. Bremond, M.Thonnat: “Online learning of activitiesfrom video”, in proc. of AVSS 12,Beijing 2012. [3] L. Patino, H. Benhadda, F.Bremond: “Data Mining in a VideoDatabase”, in “Intelligent VideoSurveillance Systems”, Wiley, OnlineLibrary, 2013,DOI:10.1002/9781118577851.ch14

Please contact:

François BremondInria, FranceE-mail: [email protected]

In smart cities, the convergence ofmobile communication, sensors andonline social networks technologies haslead to an exponential increase in thecreation and consumption of data whichcan be linked to individuals (i.e., per-sonal data). These data are consideredby the World Economic Forum as “thenew oil”, creating an unprecedentedpotential for new applications and busi-ness. However, there are privacy con-

cerns linked to the Web model currentlyin use and its underlying businesses.Frequent breaches of users' privacy arethwarting enthusiasm for the use of thisdata.

In smart cities, privacy is a fundamentalvalue, upheld by social sustainability.The personal data given back to citizenscan potentially describe all of the activi-ties undertaken in day-to-day lives. To

exploit the data, a new sharing andusage model which is more user-centricand prioritises privacy must beinvented. Our vision [1] is to launch asea change in the way data and applica-tions are managed, enabling users toexercise control over how their data isused by introducing trusted cells in thearchitecture. Trusted cells are units ofhardware and software, owned by citi-zens, which are able to perform data

trusted Cells: Ensuring Privacy for the Citizens

of Smart Cities

by Nicolas Anciaux, Philippe Bonnet, Luc Bouganim and Philippe Pucheral

The Smart City concept is founded on the collection, sharing and analysis of data that is either

about citizens or produced by them, with the view to enhancing efficiencies and the social

sustainability of cities. The current Web model which is fully centralised is not appropriate for

managing such data as it raises potential privacy abuse and misuse issues. In the Trusted Cell

project, we propose the addition of a personal dimension to the Web model: each citizen would

possess their own personal data server which would provide tangible privacy and security

guarantees and help individuals to share and disseminate their data properly.

management tasks in a privacy-pre-serving manner. Our goal is (i) to pro-vide data management and access con-trol techniques embedded into trustedcells which limits the personal dataexposed to the services and applicationsto a minimum subset; and (ii) to rely ontrusted cells to implement a usage con-trol model which protects the personaldata used by applications against unex-pected disclosures once they have leftthe security sphere of a trusted cell.

Our approach is based on the vision thatprivacy-preserving trusted cells, i.e.,trusted gateways deployed at home(fixed) or in users' hands (portable), canenforce a usage control model in con-nection with an untrusted cloud infra-structure. From a technical point ofview, the challenges associated withthis approach are:• Developing trusted cells which pro-

vide strong security guarantees: in aSmart City, many devices can be con-sidered as potential trusted cellsincluding microcontroller basedtokens (e.g., SIM cards or sensors) ormobilephones endowed with anARM Trustzone processor. Prelimi-nary proposals start addressing thesupport of relational operations likeselection, projection and join withinsecure microcontrollers [2].

• Designing usage control models, aswell as the mechanisms needed toenforce them. We are currentlyexploring a solution which relies on

two building blocks that enforceaccess and usage control modelswhen the data is transmitted outsidethe trusted cell to perform a computa-tion. We designed and implemented aset of rich operations within the trust-ed cell that can be combined toreduce the computations to be per-formed outside the trusted cells. Thenwe investigated solutions based onthe definition of sandboxed computa-tion containers, that are able to con-nect to the trusted cell and extract theraw data whilst guaranteeing thatunexpected data disclosures will notoccur. We expect that these effortswill give rise to a new paradigm forthe design of privacy preserving,user-centric applications.

• Global and anonymous computations.To compute results at a population-level, a large number of trusted cellsmay be required. The aim is to onlyreveal the results, but organize thecomputation so that the raw data andtheir owner’s identities remain hid-den. Large sets of trusted cells withlimited resources and potentially, lowconnectivity, may be involved. Thefeasibility of this approach has beenillustrated in previous work [3] whichaddressed the problem of Privacy-Preserving Data Publication in thiscontext. Computing regular queries,e.g., aggregations to discover over-crowded roads or conduct public sur-veys, is still an open issue.

Link:

CityLab@Inria Project Lab on SmartCities: http://citylab.inria.fr

References:

[1] N. Anciaux et al: “Trusted Cells: ASea Change for Personal DataServices”, in CIDR, 2013. [2] N. Anciaux et al: “MILo-DB: apersonal, secure and portable databasemachine”, in Distributed and ParallelDatabases 32(1), 2014. [3] T. Allard, B. Nguyen, P. Pucheral:“METAP: revisiting Privacy-Preserving Data Publishing usingsecure devices”, in Distributed andParallel Databases 32(2), 2014.

Please contact:

Nicolas AnciauxInria and University of Versailles,France E-mail: [email protected]



Figure 1: Personal data management

in Smart Cities using trusted cells.

Alice (A) and Bob (B), citizens of a

Smart City, are equipped with trusted

cells, which acquire data from several

sources. For example, their home is

equipped with smart meters generating

hundreds of measurements per hour.

Companies provide them with energy

management services based on smart

meter data disaggregation (i.e.,

machine learning algorithms detecting

appliance signatures and deriving their

use in time). These data are managed

into trusted cells which act as trusted

gateways that enforce a usage control

model, in connection with an untrusted

cloud infrastructure.


In a typical modern city, many differentagencies are activated, in order to pro-vide citizens with the services thatunderpin the seamless operation of theurban centres, such as the fire brigade,police department and health care serv-ices providers. The management of theactions necessary for this objective is inprinciple carried out by each organiza-tion independently, since a continuouscoordination is considered too costlyand arduous. A notable exception is thehandling of emergency incidents due totheir critical nature.

Over the last few years, many SmartCity projects have been launched withthe purpose of augmenting the capabili-ties of the relevant agencies in carryingout their tasks. For example, inAmsterdam, the Smart TrafficManagement Project is used to admin-ister traffic flow and the European

Union (EU) is preparing to launch thenew emerging trend of Connected Carswhich aims at substantially improvingambulance response times. However,the approach underlying these applica-tions is specialized rather than holistic,as is also the case with the variouspublic agencies involved.

Nonetheless, constantly maintaining acoordinated and synergistic plan can besignificantly optimized for a number ofreasons. First, this approach translatesto a substantial cost reduction in termsof human resources and infrastructure.In addition, the activities of two or moreservicing organizations are often inter-dependent: properly regulated trafficcongestion by the traffic police canfacilitate enormously ambulances toreach their destination on time. Hence, aunified approach could better serve theinterests of the stakeholders and ulti-

mately the citizens of a city.Furthermore, the experience andexpertise gained by exercising this typeof administration, might prove to becrucial in the event of a large-scaleemergency incident, such as an earth-quake or extreme weather event.

The ISL laboratory of FORTH-ICS isdeveloping the Smart City OperationCenter (SCOC), a platform for the cen-tral coordination of the various urbanservices (Figure 1 shows the mainwindow of the platform). The systemreceives requests from different loca-tions in the city for servicing tasks andorganizes the routes of the differentservice providing units (e.g. policetrucks or ambulances). This is accom-plished in such a way that the aforemen-tioned demands are satisfied in the leastamount of time possible. Figure 2 pres-ents a toy example of agents cooper-

Smart City Operation Center: A Platform

to Optimize Urban Service Rendering

by Filippos Gouidis, Theodore Patkos and Giorgos Flouris

The delivery of Public Services to citizens is becoming a challenging task, as urban populations

increase and the related networks are becoming more complex and entangled. In this article, we

present a platform designed for the optimal coordination of a city's public agencies and emergency

response teams. Such coordination is particularly relevant in emergency situations when the

infrastructures are pushed to their limits.

Figure 1: The main window of the SMOC application. In the left sub-window, the current service requests are displayed and in the right, the

optimal agent allocation required to fulfil them.

ating, in order to minimize the time ofservice. The development of the appli-cation begun in December 2013 and inits current version it can support thecoordination of small scale emergencyscenarios. The plan for the next stage ofthe development includes the extensionof the platform so that it can cope withlarger scale emergency events, such asthose that occur in mid-sized cities. Thelong-term objectives are to address thedaily administrative requirements asso-ciated with all servicing organizationsoperating in a medium to large city.

The efficient operation of SCOC relieson the optimal coordination, coopera-tion and communication between alarge number of different agents, eachof which has certain capacities and,often, oriented towards separate goals.For this reason, we opted for a synthesisof formal methods adopted from thefield of Artificial Intelligence (AI) andKnowledge Representation andReasoning. Therefore, this work can beregarded as an amalgam of featuresstemming from solutions of problemslike the multiple Traveling SalespersonProblem, graph traversal, planning withtemporal and causal constraints andtemporal reasoning. The efficient com-bination of these techniques, however,is far from trivial. The interdependen-cies between the agents, the partialobservability of the environment, thenon-deterministic nature of certainactions, the unpredictability of manyfactors, the appropriate management ofvast amounts of information data andother issues related to imperfect com-munication are some of the challengingcomplications that have to be consid-ered.

On the technical side, our implementa-tion utilizes two eminent AI paradigms:Action Languages (ALs) and AnswerSet Programming (ASP). The former isa family of logical formalisms that issuitable for representing actions andreasoning about their effects. Due totheir efficiency in dealing with proper-ties of dynamically-chaning worlds,ALs hold a pivotal role in providingsolutions, like model checking andplanning, for a multitude of dynamicdomains. In our case, they formally rep-resent the dynamic aspects of problemand allow for the formal validation andverification of the solutions generatedby the system.

ASP, on the other hand, is a logic-basedparadigm that is well suited for per-forming commonsense reasoning, formodelling an agent’s belief sets and forhandling defeasible inferences, prefer-ences and priorities. Some of ASP'smain advantages include its efficiencyin dealing with incomplete informationand the usage of a very expressive, yet,non-complex modelling language. Inaddition, ASP solvers are rapidlyimproving, rendering this declarativeparadigm to be considered a prominentcandidate for addressing large-scaleproblems. For instance, recently ASPwas applied with great success to thehighly demanding task of team buildingin the Gioia-Tauro seaport [1].However, despite their merits, each ofthe paradigms has its own drawbacksand one of the biggest challenge is tocombine them in a way that harnessesthe best of both worlds without assimi-lating the corresponding pitfalls.

In conclusion, we believe that this workoffers practical benefits and will con-tribute to the more profound compre-hension of the theoretical foundationsbehind related challenging problems.Moreover, it could serve as a guide forthe development of other Smart Cityapplications revolving around theemerging field of Multi-Agent Systems.

Link:

http://www.ics.forth.gr/isl/

Reference:

[1] Grasso, Giovanni, et al. "An ASP-based system for team-building in theGioia-Tauro seaport." Practical Aspectsof Declarative Languages. SpringerBerlin Heidelberg, 2010. 40-42

Please contact:

Filippos Gouidis, Theodore Patkos andGiorgos FlourisFORTH-ICS, GreeceTel:+30-2810-391632E-mail: [email protected];[email protected]; [email protected]



Figure 2: An example of a planned coordination generated by

the system. In order for the ambulance to reach its destination

in the least possible time (node e), a police car is dispatched in

an intermediate location (node c).


Many European cities are currentlyaiming to become Smart Cities, withmore than 1000 having already joinedthe European Smart Cities StakeholderPlatform. However, many face problemsin defining a ‘high-level inclusivevision’ that is not just driven by a fewdecision key makers. Using such a top-down process is risky, not only becauseof the risk that it might address thewrong priorities, but because it doesn’taccurately reflect the complex web ofinteractions between multiple peopleand organizations that cities are built on.A key factor of success is to activelyinvolve stakeholders and empower themas actor in the process of change. Overthe last decade, a number of new con-cepts have emerged to support this pre-ferred collaborative approach includingco-creation, user innovation and open(source) innovation [1]. They involvespecific instruments such as Living Labs[2], multidisciplinary incubators, co-working spaces and creative hubs.

It is widely recognized that ICT plays akey role in supporting the developmentof Smart Cities. Automating and opti-mizing the management of transporta-tion, energy and water requires the effi-cient monitoring and processing of data,making use of concepts such as sensornetworks, Internet-of-Things (IoT), bigdata and open data. It is essential thatICT experts are involved in the co-cre-ation process of a Smart City.

Here, we share some lessons learnt fromtwo co-innovation stories that feature astrong ICT component, currently takingplace under the eGov Wallonia initiativewith which our research center is deeplyinvolved. The co-innovation process isorganized at the regional level becausemultiple cities share similar concernsand many have common public trans-portation systems, power managementsystems and water distribution net-works.

Story 1: How open publictransportation data can be enabledThe Walloon public bus transportsystem is operated by the public TECcompany which has its own websiteoffering a few services (e.g., timetables)but leaves a number of needs unan-swered. These include questions aroundmobile access and multi-modal sched-ules. This gap led to some independentactors developing their own mobile appsusing web-scrapping techniques. TheTEC responded aggressively to thismove, a common reaction in these situa-tions. The company put pressure on theweb application developer, accusingdevelopers of data theft. Subsequently,the company realized that the transportdata belonged to the public and learntabout the open data movement. Theythen joined the eGov hackathon initia-tive, providing open access to their dataat a hackathon specifically dedicated toenhancing mobility organization in theco-working space of Wallonia’s capitalcity, Namur. This move resulted in a“TEC Real Time” project which is cur-rently exploring how best to exploit(anonymous) data from registered usersand the bus GPS information. This sce-nario illustrates how important it is tofirst establish trust and then explore newpossibilities together, many of which aremade possibly by ICT.

Story 2: How communities can bemobilized efficiently to improve theaccessibility of public infrastructuresThis issue is particularly important forpeople with reduced mobility. Changingpublic infrastructure is a long-termprocess, however, it is possible toimprove information flows regardingaccessibility issues, thus making short-term adaptations possible. Associationsall over Europe are currently fighting forthis. In Wallonia, CETIC is helping thelocal association, CAWAB, to use ICTtools to achieve a balance betweenmobilizing communities (using many

non-expert volunteers) and relying onjust a few experts (who are unable tocope with massive workloads) [3]. Thisresulted in the deployment of two com-munity websites. The first, Access-i, ismainly dedicated to informing peopleabout accessibility with a possible returnchannel. The second, CENA, is aMoodle-based open collaborative plat-form for accessibility experts, enablingnew comers to learn from the estab-lished experts.

The stories showcased in this articleillustrate the smart use of tools likehackathons, co-working spaces andcommunity platforms can yield benefitswhich make a concrete contributiontowards developing Smart Cities. Ournext step is to optimally use the LivingLabs that are currently being deployed(or will be deployed are as part of thenext FEDER program 2014-2020).

Links:

http://eu-smartcities.euhttp://hackathonegovwallonia.net http://www.access-i.behttp://cena.accessible-it.org

References:

[1] A. Kambil, G.B. Friesen, A.Sundaram: “Co-creation: A new sourceof value”, Outlook Magazine, 1999[2] B. Bergvall-Kareborn, M. Host, A.Stahlbrost: “Concept design with aliving lab approach”, in proc. of 42ndHawaii International IEEE Conferenceon System Sciences, 2009[3] C. Ponsard, V. Snoeck: “UnlockingPhysical World Accessibility throughICT: a SWOT Analysis”, ICCHP’14,2014.

Please contact:

Christophe PonsardCETIC, BelgiumE-mail: [email protected]

Building Smarter Cities

through ICt-driven Co-Innovation

by Christophe Ponsard, Robert Viseur and Jean-Christophe Deprez

The Smart City concept is actively being discussed in many places across Europe. However, turning this ideal vision into a

practical roadmap is challenging because it requires effective coordination and cooperation between the range of

organizations involved in the city operations. In this article, we discuss the use of co-innovation techniques in a number of

Belgian cities which gather people with complementary skills for collaborative projects, thus enabling a range of potential

projects to be considered. In these projects, information and communication technologies (ICT) would play a key role.


Supporting the Design

Process of Networked

Control Systems

by Alexander Hanzlik and Erwin Kristen

The increasing complexity of electronic control systems

requires new methods for design, analysis and test of

such systems. The DTFSim Data Time Flow Simulator

developed at the Austrian Institute of Technology is a

discrete-event simulation environment for model-based

design, analysis and test of networked control systems

for automotive applications.

Automotive Networked Control SystemsAn automotive control system consists of a set of devicesthat control one or more functions of the vehicle, such as anelectronic brake. Modern automotive control systems arenetworked control systems (NCS). A typical NCS containsthe following components:• Sensors to acquire information from the physical environ-

ment• Controllers to provide decision and commands• Actuators to perform the control commands• Communication network for information interchange

between sensors, controllers and actuators

In an NCS, control and feedback signals are exchangedamong the different system components in the form of infor-mation packages through the communication network.

Principle of OperationThe DTFSim is based on the idea that a NCS can be built byrepetitive use of the following parts:

Component part: Components are typically a piece of hard-ware, such as an ECU, a sensor or an actuator. Each compo-nent executes one or more functions. Components are spa-tially separated and connected via direct links (wire) or acommunication network.

Network part: The network part is responsible for communi-cation between components that are not directly linked. Thispart comprises the network architecture and the network pro-tocol, such as an Ethernet bus.

European

Research and

Innovation

Figure 1: Networked control systems containing sensors, controllers,

actuators and a communication network.

Source: http://www.hycon2.eu/


Function part: The function part contains the implementa-tion of the component functions, for instance, a brake torquecalculation algorithm.

Timing part: The timing part contains the timing propertiesof the system, such as the worst-case execution times ofcomponent functions.

The system architecture, consisting of the component partand the network part, is built using elements from a modularassembly system provided by the DTFSim. These elementsare grouped together to from more complex structures, suchas an ECU or an Ethernet bus. Typical elements are sensors,processors and actuators. Component functions (functionpart) are added to the different elements to be able to performspecific control tasks. Finally, each element has a propaga-tion delay (timing part).

DTFSim models consist of “event chains”, where an eventchain is a directed path from a sensor to an actuator, with anarbitrary number of elements in between. The simulationaims to determine essential system characteristics, such asthe bus load (the load of the communication network) and thecontrol signal latencies (the propagation times of control sig-nals from sensors to actuators). This is achieved by stimu-lating the system at the model inputs (sensors) and byobserving the event and data flow along event chains overtime to the actuators.

DTFSim WorkflowA typical workflow comprises the following steps:• Configuration: The first step is the generation of the sys-

tem configuration, which consists of the system architec-ture (component and network part) and the system behav-iour (function and timing part).

• Drive Cycle: A drive cycle consists of a set of input eventsfor each sensor of the system, e.g., the different positionsof the vehicle brake pedal during simulation. The sum ofall input events is an event list used for stimulation of thesystem.

• Simulation: Based on the configuration and the drivecycle, the simulation is executed until all events have beenprocessed. Each event that occurred during simulation isstored in the simulation results log file.

• Post-Processing and Visualization: The last step is thepost-processing of the simulation results. In this step, dataof interest is extracted from the simulation results log file.This data is then used to produce visualizations of systemcharacteristics of interest, such as the bus load over simu-lation time or the propagation time distributions of dedi-cated control signals.

Current Applications and Future WorkThe DTFSim has been developed in the course of the EUARTEMIS project POLLUX, which was related to thedesign of electronic control systems for the next generationof electric vehicles. Currently, the DTFSim is deployed intwo EU ARTEMIS projects: In MBAT (Combined Model-based Analysis and Testing of Embedded Systems), theDTFSim is deployed in an automotive used case where theperformance of an electronic braking system is analyzed bymeans of a simulation model. In CRYSTAL (Critical SystemEngineering Acceleration), the DTFSim is also deployed inan automotive use case where it is used for timing analysis ofan automatic speed limitation application.Future work relates to the integration of the DTFSim into anautomated verification and validation process for embeddedcontrol systems.

Links:

http://www.mbat-artemis.eu/home/http://www.crystal-artemis.eu/

Reference:

[1] A. Hanzlik, E. Kristen: “A Methodology for Design,Validation and Performance Analysis of Vehicle ElectronicControl Systems”, in proc. of AMAA 2013, Berlin

Please contact:

Alexander Hanzlik, Erwin Kristen, AIT / AARIT, AustriaE-mail: [email protected],[email protected]

Figure 2: The DTFSim

Data Time Flow

Simulator workflow

comprising the steps

configuration, drive

cycle, simulation, and

postprocessing and

visualization.

Figure 1: The risk of grounding was detected (yellow ring) three

minutes before it actually occurred (purple square).

Figure 2: A small fishing boat (red) repeatedly stopped in the middle

of the shipping route. It was almost hit by a large freight ship (purple).

54 ERCIM NEWS 98 July 2014

Each of these behaviours requires different anomaly detec-tors which consider different features calculated from dif-ferent parts of the data, often on different time scales. Someare suitably handled with statistical methods, others with rulebased approaches and yet others with a combination. Moststatistical anomaly detection approaches used in this studywere based on the Incremental Stream Clustering framework[1] developed by SICS, and most of the rule based detectionis based in the Situation Detector platform [2] developed bySaab.

Results The methods developed in this study were evaluated on realmaritime data collected from the Baltic sea. The success ofthis approach in detecting interesting situations is illustratedby the following examples:• A risk of running ashore - Saab has combined a rule based

detection of ships approaching shallow waters with a sta-tistical anomaly detector for ships that deviate from theusual routes in the area. The detector detected actualgroundings up to three minutes prior to such an eventoccurring (see Figure 1).

• Dangerous behaviour - the SICS’s movement patternbehaviour analysis detected a small fishing boat that was

How to Detect Suspect

Behaviour at Sea

by Anders Holst

A new model to detect and visualize ships behaving in a

strange or suspicious manner can help authorities to

prevent or mitigate accidents and detect illegal

activities at sea.

The three-year SADV project, which finished in 2013, hasbeen a collaboration between SICS Swedish ICT, Saab AB,the Swedish Coast Guard, the Swedish Customs Service, theSwedish Armed Forces and the Swedish Space Corporation.The aim of the project has been to support the maritime sur-veillance operators by providing them with an increased situ-ation awareness. This is done using automatic tools whichdetect vessels bahaving suspiciously or uncharacteristically.Increasing volumes of maritime traffic, combined withthreats from criminal activities and environmental issues,just to name a few, there is a high demand for tools that cansupport the real-time analysis of the huge amounts of mar-itime situation data that are produced.

Three approaches for anomaly detection There are essentially three different approaches to detectanomalies and the framework developed in this study sup-ports all three:1. statistical anomaly detection, where a statistical model is

built on normal situations and situations that are veryunlikely to come from that model are considered anom-alies;

2. rule based anomaly detection, where rules are designed todetect situations of interest; and

3. model based (or simulator based) anomaly detection,where real observations are compared to simulated results(i.e., what would have happened in a normal situation). Adifference indicates an anomaly.

While it is easily seen that approaches two and three requireextensive expert involvement to design the rules and simu-lator, approach one would appear to require no expertinvolvement. However, this approach still requires extensivedomain knowledge, first to determine what kind of anom-alies would be interesting to detect and secondly, to ensurethat the model considers the features that are relevant fordetecting those anomalies. Typically, the data requires someprocessing as these features are not directly represented inthe raw data.

Uncharacteristic behaviours consideredA number of uncharacteristic vessel behaviours have beenconsidered in this study. They are: • unusual speeds and directions (e.g., going the wrong

direction in a route),• unusual movement patterns (e.g., too many stops or turns

at-sea),• making a rendezvous at sea.• unusual route choices (as compared with other ships), and• data inconsistencies (e.g., a ship changing identity or

using the same identity as another ship).

Research and Innovation

Figure 3: The architecture of the anomaly detection module

55

If the indications from one anomaly detector can be used asan input feature to another anomaly detector, it is possible tocombine the rule based and statistical detectors in a moreadvanced ways, as opposed to just having them run in par-allel in the same system [3]; statistical detections can bereferred to in a rule, or the number of times that a rulematches can be considered in a statistical detector.

ConclusionThe SADV project has significantly extended previousmethods used for anomaly detection within the maritime sur-veillance sphere and shown that these methods can detectinteresting and relevant situations.

Link:

http://www.sics.se/projects/sadv-statistical-anomaly-detec-tion-and-visualization-for-maritime-domain-awareness

References:

[1] A. Holst and J. Ekman: “Incremental stream clusteringfor anomaly detection and classification”, in EleventhScandinavian Conference on Artificial Intelligence, SCAI2011[2] J. Edlund, M. Grönkvist, A. Lingvall and E. Sviestins:“Rule-based situation assessment for sea surveillance”, inproc. of SPIE Defence and Security Symposium 2006[3] A. Holst et al.: “A Joint Statistical and Symbolic Anom-aly Detection System: Increasing performance in maritimesurveillance”, in FUSION 2012.

Please contact:

Anders HolstSICS Swedish ICTTel: +46 (0)8 633 1593E-mail: [email protected]

repeatedly stopping in the middle of a shipping route andmoving irregularly in the wrong direction (apparently fish-ing). It is almost hit by a large freight ship that managed toveer at the last minute (Figure 2).

• Several ships using the same identity - the SICS’s incon-sistency detector identified several ships using the wrongidentity number. These are mostly genuine mistakes; how-ever, there are examples of ships in other parts of theworld intentionally changing their identity numbers to cir-cumvent trading embargoes.

• Meetings between ships at sea - most meetings, unlessthey are between certain ship types (i.e., coast guard, res-cue vessels, etc.), may potentially indicate illegal activitiessuch as smuggling or possibly trawling in prohibitedareas.

How the model worksThe task of combining all three different types of anomalydetection in one framework was not trivial, but we have man-aged to design a general architecture that fits with severalexisting maritime surveillance platforms and is also able tohandle several anomaly detectors of different kinds. Thearchitecture of the anomaly detection module is schemati-cally shown in Figure 3.

The key to this architecture is identifying information flowsthat are common across all the surveillance platforms and forall kinds of anomaly detectors. All detectors need situationdata from the platform, and they need to provide anomalyindications back, based on that data. They also need to sup-port user configuration and inspection.

The next step is to identify similarities between the differentkinds of anomaly detectors. They all need to extract the rele-vant features from the provided data, use their respectivemethods to determine whether those features represent ananomaly, and then finally transform that judgement into an“indication” which is complete with details of what, whenand why.


Research and Innovation

A Breakthrough

for Balanced Graph

Partitioning

by Fatemeh Rahimian

Researchers at SICS have invented a new way to

partition graphs into a given number of clusters, such

that the related information belongs to the same cluster

and the number of connections between clusters are

minimized. Graph partitioning, a well known NP-

Complete problem, has been already thoroughly

investigated for small and medium sized graphs, but

this distributed algorithm works on very large graphs.

Our world is increasingly connected, and this is changing ourlives in ways that we still do not fully understand. New toolsand technologies have given us the unprecedented potentialto make sense of the huge inter-connected datasets that existin various fields of science, from social networks to biolog-ical networks. Relations can be mathematically described as“graphs”; for example, a network of friends on Facebook canbe described by a graph, which shows a user’s closest friendsand who they are related to. The increasing size of datasetsmeans that graph sizes are also increasing: this means thatthey must be partitioned into smaller clusters so they can bemore easily managed on multiple machines in a distributedfashion.

The new solution for this is a distributed heuristic basedalgorithm that can efficiently partition big graphs into agiven number of clusters of equal size or any given size. Thekey point in this approach is that no global knowledge isrequired, and simultaneous access to the entire graph is notassumed.

The algorithm used to solve the problem is very intuitive. Itstarts by randomly assigning vertices to partitions. Over thecourse of the algorithm, vertices from different partitionsiteratively negotiate with each other and exchange their par-tition assignments, if that exchange results in a better parti-tioning [1]. When this iterative optimization converges,neighboring vertices of the graph will mostly end up in thesame partition, while there will be very few edges that crossdifferent partitions (See Figure 1). The algorithm is inher-ently a local search optimization enhanced using a simulatedannealing technique. A variant of this work has been specifi-cally designed for graphs with power-law degree distribu-tion, where vertex-cut partitioning has proved more efficientthan edge-cut partitioning [2]. Both solutions are kept assimple as possible, so that the algorithm can be easilyadopted by various real world applications.

With the recent advances in graph database technologies, it isanticipated that leading companies in this field will soonmove to make use of graph partitioning for scaling up.Successful partitioning is an important ingredient in efficientbig data analysis and that is why this solution has attracted alot of attention in the community.


Figure 1: 4elt sparse graph, partitioned into 4 balanced components

that are indicated by colors. Very few links are on the border of

regions with different colors. For her achievement [1] Fatemeh

Rahimian received the Best Paper Award at the 7th IEEE

International Conference on Self-Adaptive and Self-Organizing

Systems in Philadelphia in September 2013. Moreover, in May 2014

she obtained a doctoral degree that was partly based on this work.

Link:

http://e2e-clouds.org/

References:

[1] Rahimian, Fatemeh, Amir H. Payberah, Sarunas Girdzi-jauskas, Mark Jelasity, and Seif Haridi. "Ja-be-ja: A distrib-uted algorithm for balanced graph partitioning." In the 7thIEEE Conference on Self-Adaptive and Self-OrganizingSystems (SASO), 2013.[2] Rahimian, Fatemeh, Amir H. Payberah, Sarunas Girdzi-jauskas, and Seif Haridi. "Distributed vertex-cut partition-ing." In the 14th IFIP International conference on Distrib-uted Applications and Interoperable Systems (DAIS), 2014.

Please contact:

Fatemeh RahimianSICS Swedish ICTE-mail: [email protected]


Call for Participation

fMICS’14 - 19th

International

Workshop on formal

Methods for Industrial

Critical Systems

Florence, Italy ,11-12 September 2014

The aim of the FMICS workshop seriesis to provide a forum for researchers whoare interested in the development andapplication of formal methods inindustry. In particular, FMICS bringstogether scientists and engineers that areactive in the area of formal methods andinterested in exchanging their experi-ences in the industrial usage of thesemethods.

Topics of interest include (but are notlimited to):• design, specification, code generation

and testing based on formal methods• methods, techniques and tools to sup-

port automated analysis, certification,debugging, learning, optimization andtransformation of complex, distrib-uted, real-time systems and embeddedsystems

• verification and validation methodsthat address shortcomings of existingmethods with respect to their industri-al applicability (e.g., scalability andusability issues)

• tools for the development of formaldesign descriptions

• case studies and experience reports onindustrial applications of formalmethods, focusing on lessons learnedor identification of new researchdirections

• impact of the adoption of formalmethods on the development processand associated costs

• application of formal methods in stan-dardization and industrial forums.

FMICS’14 is co-located with QEST2014, SAFECOMP 2014, EPEW 2014,and FORMATS 2014 in the frame ofFLORENCE 2014 - a scientific week oncomputer safety, reliability, perfor-mance, and quantitative analysis.

More information:

http://fmics2014.unifi.it/http://www.florence2014.org/

sions” that will run in parallel during thethree days of the conference.

TutorialsTutorials will be given on Friday 5December 2014. The number of partici-pants to the tutorials is limited andrestricted only to those who attend theconference.

PublicationsThe journal Computational Statistics &Data Analysis will publish selectedpapers in special peer-reviewed or reg-ular issues.

More information:

http://www.cmstatistics.org/ERCIM2014/


7th International

Conference of the

ERCIM Working Group

on Computational

and Methodological

Statistics

Pisa, Italy, 6-8 December 2014.

The 7th International Conference onComputational and MethodologicalStatistics (ERCIM 2014) is organized bythe ERCIM Working Group onComputational and MethodologicalStatistics (CMStatistics) and theUniversity of Pisa. The conference isheld jointly with the 8th InternationalConference on Computational andFinancial Econometrics (CFE 2014).The conference has a high reputation ofquality presentations. The last editionsof the joint conference CFE-ERCIMgathered over 1200 participants.

Aims and ScopeAll topics within the Aims and Scope ofthe ERCIM Working GroupCMStatistics will be considered for oraland poster presentation.

Topics include, but are not limited to:robust methods, statistical algorithmsand software, high-dimensional dataanalysis, statistics for imprecise data,extreme value modeling, quantileregression and semiparametric methods,model validation, functional dataanalysis, Bayesian methods, optimiza-tion heuristics in estimation and model-ling, computational econometrics, quan-titative finance, statistical signal extrac-tion and filtering, small area estimation,latent variable and structural equationmodels, mixture models, matrix compu-tations in statistics, time series modelingand computation, optimal design algo-rithms and computational statistics forclinical research.

SessionsThe joint conference CFE-ERCIM willhave five plenary zessions. Moreover,ERCIM 2014 will have four “specialinvited sessions” and a significantnumber of “organized invited sessions”on key topics, and “contributed ses-

57

Events


SAfECOIM 2014 -

33rd International

Conference on

Computer Safety,

Reliability and

Security

Florence 10-12 September 2014

SAFECOMP is an annual event cov-ering the state-of-the-art, experience andnew trends in the areas of safety, secu-rity and reliability of critical computerapplications. SAFECOMP providesample opportunity to exchange insightsand experience on emerging methods,approaches and prac tical solutions. It isa one-stream conference without par-allel sessions, allowing easy net-working.

SAFECOMP 2014 will be part of FLO-RENCE 2014, a one-week scientificevent with conferences and workshopsin the areas of formal and quantitativeanalysis of systems, performance engi-neering, computer safety, and industrialcritical applications.

More information:

http://www.safecomp.orghttp://www.ewics.orghttp://www.florence2014.org

58 ERCIM NEWS 98 July 2014

• formal and semi-formal techniquesfor verification and validation

• experimental evaluations of resilientsystems

• quantitative approaches to ensuringresilience

• resilience prediction• case studies & applications• empirical studies in the domain of

resilient systems• cloud computing and resilient service

provisioning• resilient cyber-physical systems and

infrastructures• global aspects of resilience engineer-

ing: education, training and coopera-tion.

The workshop is organised by theERCIM Working Group SERENE.

The proceedings of SERENE 2014 willbe published as a volume in SpringerLecture Notes in Computer Science(LNCS).

Autumn SchoolAn autumn school will be held rightbefore the SERENE workshop. Theautumn school will explore theresiliency of cyber physical systems.

More information:

http://serene.disim.univaq.it/


SERENE’14 -

6th International

Workshop on

Software Engineering

for Resilient Systems

Budapest, 15-16 October 2014

Unprecedented level of complexity ofmodern software and software-basedsystems makes it difficult to ensure theirresilience - an ability of the system topersistently deliver its services in adependable way even when facingchanges, unforeseen failures and intru-sions. Yet we are observing the increas-ingly pervasive use of software inevolvable and critical systems liketransportation, health care, manufac-turing, and IT infrastructures. This trendurges the research community todevelop powerful methods for assuringresilience of software-intensive sys-tems.

These challenges have also appeared inthe scope of the current Horizon 2020calls that aim at developing tools andmethods for incorporating resilienceinto evolving software systems; andalso in calls related to specific applica-tion areas like advanced cloud infra-structures and services, smart objects,and robotics.

The SERENE 2014 workshop providesa forum for researchers and practi-tioners to exchange ideas on advancesin all areas relevant to software engi-neering for resilient systems, including,but not limited to:• design of resilient systems• requirements engineering & re-engi-

neering for resilience• frameworks, patterns and software

architectures for resilience• Engineering of self-healing autonom-

ic systems;• design of trustworthy and intrusion-

safe systems• resilience at run-time (mechanisms,

reasoning and adaptation)• verification, validation and evalua-

tion of resilience• modelling and model based analysis

of resilience properties;

Call for Papers

International

Workshop on

Computational

Intelligence

for Multimedia

Understanding

Paris, 1-2 November 2014

Multimedia understanding is an impor-tant part of many intelligent applicationsin our social life, be it in our households,or in commercial, industrial, service,and scientific environments.

Analyzing raw data to provide themwith semantics is essential to exploittheir full potential and help us managingour everyday tasks. Our purpose is toprovide an international forum topresent and discuss current trends andfuture directions in computational intel-ligence for multimedia understanding.The workshop, organized by the ERCIMMUSCLE Working Group, also aims atfostering the creation of a permanentnetwork of scientists and practitionersfor an easy and immediate access topeople, data and ideas.

Scope and TopicsTopics include, but are not limited to:• multimedia labeling, semantic annota-

tion• multimodal and cross-modal data

analysis, clustering• big data, distributed digital libraries• activity and object detection and

recognition• text and speech recognition• multimodal indexing and searching in

very large data-bases• visual data processing techniques• medical image segmentation and

model-based representation• microscopic image feature extraction

and classification

Important dates• Paper Submission: 25 July, 2014 • Notification: 12 September, 2014• Final Regular Paper Submission: 26

September, 2014

More information:

http://iwcim.isep.fr

Events

59ERCIM NEWS 98 July 2014

Roberto Scopigno

receives the

Eurographics

“Distinguished

Career Award”

2014

Roberto Scopigno is the recipient of the Eurographics

"Distinguished Career Award" 2014, an award given every

other year to a professional in computer graphics who has

made outstanding technical contributions to the field and has

shaped computer graphics in Europe.

Roberto Scopigno’s work has had a profound impact on the

field of visual computing. Of particular importance has been

his work on surface simplification, LOD and multiresolution

representations for surfaces and volumes. His papers on

these topics have been cited widely and stimulated consider-

able follow-up work. The BDAM algorithm and its 3D

extension, Tetra-Puzzles, grew over ingenious novel data

structures for hierarchical seamless space subdivision and

have inspired new multiresolution applications for the

inspection of terrain models and gigantic meshes. He is a

world-wide leader in the development of novel algorithms

and techniques for Cultural Heritage and for the acquisition,

preservation and visualization of digital copies of physical

artifacts inherited from the past.

Roberto Scopigno is a recognized scientific leader. He has

created a very successful research group at CNR Pisa, and

several of his former students have themselves become well-

known productive researchers. He has demonstrated a

strong leadership in Computer Graphics research in Europe,

and he has been a major actor in shaping Computer Graphics

in Italy. Roberto has significantly advanced the field through

his work and energy, and by setting a personal example.

More information:

The complete award citation is available at:

https://www.eg.org/index.php/component/content/article/57

-awards/distinguished-career-award-recipients/340-

scopigno-2014

https://www.eg.org/index.php/awards/career-award

European Project

on Precision Farming

Increasingly, remote sensing data is made publicly available

by organizations like ESA and NASA. Based on these data,

and in combination with other sources, new valuable appli-

cations can be created. In the European research project

Linked Open Earth Observation Data for Precision Farming

(LEO) researchers from CWI and the University of Athens

join their forces with industry partners to develop an appli-

cation for precision farming. The new application is based

on remote sensing and geospatial data. With precision

farming - advanced agriculture using GPS, satellite observa-

tions and tractors with on-board computers - the farming

process is performed as accurately and efficiently as pos-

sible. This is achieved by combining data from earth obser-

vations with other geospatial sources such as cadastral data,

data on soil quality, vegetation and protected areas. This

enables farmers to find the optimal trade-off in maximizing

yield while minimizing fertilizers and pesticides. At CWI,

the researchers will combine various data sources, transfer

them into RDF format and publish them as Linked Open

Data.

More information: http://www.linkedeodata.eu

International

Innovation Award

for Martin Kersten

Martin Kersten, Research Fellow at

CWI, is awarded the 2014 SIGMOD

Edgar F. Codd Innovations Award,

the most prestigious prize for researchers who have made

innovative and significant contributions to database systems

and databases. Kersten received the award on 26 June at the

annual ACM SIGMOD/PODS Conference. The Awards

Committee recognized the influential contributions of

Kersten to the big data problem, his outstanding achieve-

ments in scientific research in advanced database architec-

tures and his pioneering work in the realization of

MonetDB.

According to Kersten, massive accumulation of data

requires fundamental research and changes in database man-

agement systems. In MonetDB he has pioneered the

columns-store technology since 1993. Nowadays it is the

most widespread open-source column-store database man-

agement system, which is worldwide used in more than 130

countries with over 300,000 downloads. Since 2011,

column-store technology as pioneered in MonetDB has

found its way into the product offerings of all major com-

mercial database vendors in relational database systems.

More information:

http://www.cwi.nl/news/2014/international-innovation-

award-big-data-research-martin-kersten

In Brief

http://www.cwi.nl/news/2014/international-innovation-award-big-data-research-martin-kersten

https://www.eg.org/index.php/component/content/article/57-awards/distinguished-career-award-recipients/340-scopigno-2014

ERCIM is the European Host of the World Wide Web Consortium.

Institut National de Recherche en Informatique et en AutomatiqueB.P. 105, F-78153 Le Chesnay, Francehttp://www.inria.fr/

Technical Research Centre of FinlandPO Box 1000FIN-02044 VTT, Finlandhttp://www.vtt.fi/

Austrian Association for Research in ITc/o Österreichische Computer GesellschaftWollzeile 1-3, A-1010 Wien, Austriahttp://www.aarit.at/

Norwegian University of Science and Technology Faculty of Information Technology, Mathematics and Electri-cal Engineering, N 7491 Trondheim, Norwayhttp://www.ntnu.no/

Universty of WarsawFaculty of Mathematics, Informatics and MechanicsBanacha 2, 02-097 Warsaw, Polandhttp://www.mimuw.edu.pl/

Consiglio Nazionale delle Ricerche, ISTI-CNRArea della Ricerca CNR di Pisa, Via G. Moruzzi 1, 56124 Pisa, Italyhttp://www.isti.cnr.it/

Centrum Wiskunde & InformaticaScience Park 123, NL-1098 XG Amsterdam, The Netherlandshttp://www.cwi.nl/

Foundation for Research and Technology - HellasInstitute of Computer ScienceP.O. Box 1385, GR-71110 Heraklion, Crete, Greecehttp://www.ics.forth.gr/FORTH

Fonds National de la Recherche6, rue Antoine de Saint-Exupéry, B.P. 1777L-1017 Luxembourg-Kirchberghttp://www.fnr.lu/

FWOEgmontstraat 5B-1000 Brussels, Belgiumhttp://www.fwo.be/

F.R.S.-FNRSrue d’Egmont 5B-1000 Brussels, Belgiumhttp://www.fnrs.be/

Fraunhofer ICT GroupAnna-Louisa-Karsch-Str. 210178 Berlin, Germanyhttp://www.iuk.fraunhofer.de/

SICS Swedish ICTBox 1263, SE-164 29 Kista, Swedenhttp://www.sics.se/

University of GenevaCentre Universitaire d'Informatique Battelle Bat. A, 7 rte de Drize, CH-1227 Carougehttp://cui.unige.ch

Magyar Tudományos AkadémiaSzámítástechnikai és Automatizálási Kutató IntézetP.O. Box 63, H-1518 Budapest, Hungaryhttp://www.sztaki.hu/

University of CyprusP.O. Box 205371678 Nicosia, Cyprushttp://www.cs.ucy.ac.cy/

Spanish Research Consortium for Informatics and MathematicsD3301, Facultad de Informática, Universidad Politécnica de Madrid28660 Boadilla del Monte, Madrid, Spain,http://www.sparcim.es/

Science and Technology Facilities CouncilRutherford Appleton LaboratoryChilton, Didcot, Oxfordshire OX11 0QX, United Kingdomhttp://www.scitech.ac.uk/

Czech Research Consortium for Informatics and MathematicsFI MU, Botanicka 68a, CZ-602 00 Brno, Czech Republichttp://www.utia.cas.cz/CRCIM/home.html

Subscribe to ERCIM News and order back copies at http://ercim-news.ercim.eu/

ERCIM - the European Research Consortium for Informatics and Mathematics is an organisa-

tion dedicated to the advancement of European research and development, in information

technology and applied mathematics. Its member institutions aim to foster collaborative work

within the European research community and to increase co-operation with European industry.

Portuguese ERCIM Groupingc/o INESC Porto, Campus da FEUP, Rua Dr. Roberto Frias, nº 378,4200-465 Porto, Portugal

I.S.I. - Industrial Systems InstitutePatras Science Park buildingPlatani, Patras, Greece, GR-26504 http://www.isi.gr/

Universty of WroclawInstitute of Computer ScienceJoliot-Curie 15, 50–383 Wroclaw, Polandhttp://www.ii.uni.wroc.pl/

University of SouthamptonUniversity RoadSouthampton SO17 1BJ, United Kingdomhttp://www.southampton.ac.uk/

ERCIM News 98

Documents

Transcript of ERCIM News 98