Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
ERCIM
www.ercim.eu
NEWS
Special theme
The Internet
of Things
and
The Web
of Things
Also in this issue:
Keynote:
by Cees Links, Founder & CEO
GreenPeak Technologies
Editorial Board:
Central editor:
Peter Kunz, ERCIM office (peter.kunz@ercim.eu)
Local Editors:
Austria: Erwin Schoitsch, (erwin.schoitsch@ait.ac.at)
Belgium:Benoît Michel (benoit.michel@uclouvain.be)
Cyprus: Ioannis Krikidis (krikidis.ioannis@ucy.ac.cy)
Czech Republic:Michal Haindl (haindl@utia.cas.cz)
France: Thierry Priol (thierry.priol@inria.fr)
Germany: Michael Krapp (michael.krapp@scai.fraunhofer.de)
Greece: Eleni Orphanoudakis (eleni@ics.forth.gr),
Artemios Voyiatzis (bogart@isi.gr)
Hungary: Erzsébet Csuhaj-Varjú (csuhaj@inf.elte.hu)
Italy: Carol Peters (carol.peters@isti.cnr.it)
Luxembourg: Thomas Tamisier (thomas.tamisier@list.lu)
Norway: Poul Heegaard (poul.heegaard@item.ntnu.no)
Poland: Hung Son Nguyen (son@mimuw.edu.pl)
Portugal: Joaquim Jorge (jorgej@tecnico.ulisboa.pt)
Spain: Silvia Abrahão (sabrahao@dsic.upv.es)
Sweden: Kersti Hedman (kersti@sics.se)
Switzerland: Harry Rudin (hrudin@smile.ch)
The Netherlands: Annette Kik (Annette.Kik@cwi.nl)
W3C: Marie-Claire Forgue (mcf@w3.org)
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 peter.kunz@ercim.eu
Subscription
Subscribe to ERCIM News by sending an email to
en-subscriptions@ercim.eu or by filling out the form at the ERCIM
News website: http://ercim-news.ercim.eu/
Next issue
July 2015, Special theme: Trustworthy Systems of Systems
The Internet of Things is creating a new locks, the majority of which are not cur-
world, a quantifiable and measureable rently interconnected. They are con-
world, where people and businesses can nected – but not to each other. The Cees Links, Founder & CEO GreenPeak
manage their assets in better informed weather station does not provide infor- Technologies, www.greenpeak.com
ways, and can make more timely and mation to the thermostat about the cli-
better informed decisions about what mate outside. The security system is not
they want or need to do. This new con- connected to the indoor motion sensors, industry. Picture the first cars hitting the
nected world brings with it fundamental nor to the electronic door locks (it does road: there were no freeways, no road
changes to society and to consumers. By not automatically lock the forgotten back signs, no rules, no driving licences.
sensing our surrounding environment, door when the inhabitants go out). In the Pedestrians did not know to get out of the
the IoT will create many practical future, all these systems will be intercon- way. Drivers did not know how to take
improvements in our world, increasing nected, providing information to each turns at intersections. Neither drivers nor
our convenience, health and safety, while other, and reacting accordingly. We are pedestrians understood the risks and lia-
at the same time improving energy effi- currently in an emerging state of the IoT, bilities, giving no consideration to lia-
ciency and comfort. The IoT will be a with individual vertical applications that bility and insurance.
new source of wealth creation. operate as islands, and serve independent
applications (such as security alarms, We are currently at a similar stage with
IoT devices can be classified in three cate- door locking, etc.). However, the real the IoT. Just as it took decades before all
gories: (1) wearables, (2) smart home IoT will emerge when these applications required infrastructure was in place
devices, and (3) M2M devices. The first cooperate, working together, and begin around motor vehicles, it will take quite
two categories are the most important for to use each other’s ’awareness’. That is some time before it is in place around the
consumers. ‘Wearables’ are the devices when the true IoT avalanche will start. Internet. Once I led the engineering team
that people carry with them, which usu- that successfully brought WiFi to the
ally connect via Bluetooth to a smart The key component of the IoT - whether Mac laptop for Steve Jobs. After his team
phone, and from there to the Internet. wearables or smart home devices - is not had achieved this, all others followed.
This category includes devices such as the sensor, but the application. Connecting Now WiFi is everywhere - but this took
smart watches, fitness bands and devices the sensors is difficult, but extracting infor- time and work. Similarly, in the context
to help people to live more ’mindfully’ – mation from data is the essence. Useful of the Internet and Internet of Things,
increasing the wearer’s awareness of how information extracted from the data can there is a growing awareness that we
well they sleep, how much they move coach people by reaffirming when things need rules, training, legislation and
around, monitoring their vital signs, etc. go as planned or by alerting or taking enforcement. We are just starting to learn
action if something goes wrong; and data what might be needed.
Smart home devices are also part of the analytics can be used to compare situa-
IoT and usually connect to the Internet tions, to coach and to provide feedback to The IoT will change the world in an even
via ZigBee low power wireless commu- help make improvements. This is slowly more profound way than has the Internet.
nication and the home router. These starting to dawn on manufacturers and If we ask our children today how the
include all domestic devices, from lights service providers alike. People are inter- world existed before Internet, they are
and light switches to motion sensors, ested in the IoT if it helps them to improve speechless. They have no comprehen-
thermostats, door locks and automated aspects of their lives. Improvements are sion of how people could communicate
curtains. Via its WiFi connection to the not achieved by sensors alone: a com- or even live their lives without the
router, the smart phone also becomes an pletely different way of thinking is common place tools we have today. The
online dashboard and control device for required, and it will take some time for the same will happen with the IoT.
Smart Home applications. new paradigm to be fully embraced.
A decade from now, we will be
The third category, M2M (Machine to Privacy and security are key, together dependent on the knowledge derived
Machine) devices, comprises devices with data ownership. Note, these are not from the continuous stream of data from
that are directly connected to the cellular IoT issues, but general internet issues that our wearables and our smart home
network, such as cars that can report are amplified by the growth of new appli- devices, and we will have no idea how
their location (in case of an accident or cations. These issues already exist for the we managed the world and our lives
theft), or vending machines that can call internet of people, and industry and gov- before. We will be able to make better
in when their stocks are running low. ernment bodies are slowly starting to rec- informed, more accurate and more
ognize them and take action. timely decisions; and decisions that will
Many households and businesses have improve our lives, save us money, and
thermostats, weather stations, smart The growth of the IoT can be compared may even save our planet. The IoT will
lighting, security and electronic door with the growth of the automobile make the difference.
3 The Internet of Things The special theme section “The Internet of Things and
Will Change our World the Web of Things” has been coordinated by Dave
by Cees Links, Founder & CEO GreenPeak Raggett, W3C and Emmanuel Baccelli, Inria.
Technologies
Introduction to the Special Theme
8 The Promise of the Internet of Things and the
JOINt ERCIM ACtIONs Web of Things
by Emmanuel Baccelli and Dave Raggett
6 ERCIM “Alain Bensoussan” Fellowship
Programme IoT Operating Systems & Network Stacks
11 Open{WSN|Mote}: Open-Source Industrial IoT
6 Cor Baayen Award 2015 by Thomas Watteyne, Xavier Vilajosana, and Pere
Tuset
6 W3C Europe@20 Anniversary Event
12 RIOT and the Evolution of IoT Operating
7 The Role of ERCIM EEIG Systems and Applications
by Antoine Petit by Emmanuel Baccelli, Oliver Hahm, Hauke
Petersen and Kaspar Schleiser
IoT Security
17 Security and Privacy in the Internet of Things
by Ludwig Seitz
26 A Mobile End-User Development Environment 44 Mesh Joinery: A Method for Building Fabricable
for IoT Applications Exploiting the Puzzle Structures
Metaphor by Paolo Cignoni, Nico Pietroni, Luigi Malomo and
by Josè Danado and Fabio Paternò Roberto Scopigno
28 3D Web Visualization for Real-Time 46 Icing Detection and Protection for Small Unmanned
Maintenance of Smart Buildings Aircrafts
by Daniel Gaston, Christophe Joubert and Miguel by Andrea Cristofaro and Tor Arne Johansen
Montesinos
30 COMPOSE: An Open Source Cloud-Based 47 LEARNMINER – Advanced Analytics for eLearning
Scalable IoT Services Platform by Christian Bauckhage, Markus Ketterl and Ulrich Nütten
by Dave Raggett
48 Online Semantic Analysis over Big Network Data
IoT Architecture by Chengchen Hu and Yuming Jiang
31 An Integration Gateway for Sensing Devices in
Smart Environments 49 From Cooling Integrated Circuits to Efficient
by Michele Girolami, Francesco Furfari and Electrical Power Generation
Stefano Chessa by Stephan Paredes, Brian Burg, Patrick Ruch, Ingmar
Meijer and Bruno Michel
33 Fighting Networking Heterogeneity in the
Internet of Things 51 VAVID– Handling Big Data
by Elias Z. Tragos, Vangelis Angelakis and
Stefanos Papadakis
EVENTS, IN BRIEF
34 Tectons: Towards a Generalised Approach to
Programming Systems of Systems Announcements
by Geoff Coulson, Andreas Mauthe and Markus 52 AITA 2015 - Advanced Infrared Technology and
Tauber Applications
36 BETaaS: Building the Future Platform for 52 IC3K - 7th Joint Conference on Knowledge Discovery,
Development and Execution of Machine-to- Knowledge Engineering and Knowledge Management
Machine Applications
by Carlo Vallati, Enzo Mingozzi and Giacomo 52 Information Security Conference - ISC 2015
Tanganelli
53 ICEC 2015 - 14th International Conference on
IoT Use Cases Entertainment
37 Pro-IoT: Improving the Productivity of
Companies by Using the IoT 53 7th International Workshop on Software Engineering
by Heikki Ailisto, Nadine Pesonen, Pertti for Resilient Systems
Raatikainen and Jonathan Ouoba
54 PhD Positions Available in the BigStorage European
39 MusicBricks: Connecting Digital Creators to the Training Network
Internet of Music Things
by Thomas Lidy, Alexander Schindler and Michela 54 1955-2015: ITC Celebrates its 60th Anniversary!
Magas
54 SERENE 2015 - 7th International Workshop on
41 Deploying an NFC-based Inventory System: A Software Engineering for Resilient Systems
Case Study
by Patrick Poullie, Thomas Bocek and Burkhard In Brief
Stiller 55 25 Years of Python at CWI
42 TrakChain Estimates Costs for Track and Trace 55 MOOC on “Binaural Hearing for Robots”
in the Internet of (many) Things
by Miguel L. Pardal and José Alves Marques 55 Java Bug Fixed with Formal Methods
W3C Europe@20
Anniversary Event
W3C Europe will celebrate its twentieth anniversary with
a symposium on Tuesday, 5 May in the Paris town hall.
ERCIM offers fellowships for PhD holders from all W3C will celebrate 20 years of work accomplished by
over the world. European stakeholders for the benefit of the Web with a
symposium on 5 May. Speakers will include, among others,
Topics cover most disciplines in Computer Science, representatives from Paris city council, the European
Information Technology, and Applied Mathematics. Commission and the French government. The closing talk
will be given by Tim Berners-Lee, Web Inventor and W3C
Fellowships are of 12-month duration, spent in one Director. About 250 invited guests - global strategists, busi-
ERCIM member institute. Fellowships are proposed ness leaders and developers - are expected to attend.
according to the needs of the member institutes and the
available funding. More information:
http://www.w3.org/20/Europe/
Conditions
Applicants must:
• have obtained a PhD degree during the last 8 years
(prior to the application deadline) or be in the last
year of the thesis work with an outstanding academic Call for Nominations
record
• be fluent in English Cor Baayen Award 2015
• be discharged or get deferment from military service
• have completed the PhD before starting the grant. The Cor Baayen Award is awarded each year to a promising
young researcher in computer science and applied mathe-
In order to encourage mobility: matics.
• a member institute will not be eligible to host a candi-
date of the same nationality. The award consists of 5000 Euro together with an award cer-
• a candidate cannot be hosted by a member institute, if tificate. The selected fellow will be invited to the ERCIM
by the start of the fellowship, he or she has already meetings in autumn. A short article on the winner, together
been working for this institute (including phd or post- with the list of all candidates nominated, will be published in
doc studies) for a total of 6 months or more, during ERCIM News.
the last 3 years.
Nominees must have carried out their work in one of the
The fellows are appointed for 12 months either by a 'ERCIM countries': Austria, Belgium, Cyprus, Czech
stipend (an agreement for a research training Republic, Finland, France, Germany, Greece, Hungary, Italy,
programme) or a working contract. The type of contract Luxembourg, Norway, Poland, Portugal, Spain, Sweden,
and the monthly allowance (for stipends) or salary (for Switzerland, The Netherlands and the United Kingdom.
working contracts) depend on the hosting institute. Nominees must have been awarded their PhD (or equivalent)
after 30 April 2011.
Application deadlines
30 April and 30 September The award was created in 1995 to honour the first ERCIM
President.
More information and application form:
http://fellowship.ercim.eu/ Detailed information and online nomination form:
http://www.ercim.eu/activity/cor-baayen-award
Google’s ‘Physical Web’ (Google, 2014) is about beacons consumers can be sure of the other party’s identity. Identity
that broadcast web addresses to devices in their neighbour- verification is about linking identities to real world proper-
hood. This can be compared to walking down a crowded ties, e.g., the physical location of a sensor, or the full name
market street with all the store holders shouting out their and postal address of a human. There is a need for trusted
wares and special offers. This calls for personal agents or agents that provide identity verification services. Trust is
avatars that are aware of your current interests and are able to also important in relation to decisions about whether to use
recognise which beacons are relevant and which can be services: are they safe, do they come from a bone fide
safely ignored. The agent could notify you directly or could source, will they safeguard my privacy and so forth. This
perform tasks on your behalf. provides opportunities for agents that perform security and
privacy audits of services. This can be complemented by
Avatars are also related to the concept of personal zones as crowd sourced reputations and reviews. Recommendation
explored in the EU FP7 webinos project. Your personal zone systems can further provide suggestions based upon what
is an abstraction that groups all of your personal devices and other people have looked at in similar contexts.
services. It provides an overlay model for secure messaging
between your devices as well as offering zone wide services We have a lot to do. The papers in this special issue of
to trusted applications running on your devices. Your per- ERCIM News provide a small sample of the research work
sonal zone also acts as an avatar on your behalf offering serv- that is helping to realise the potential for connected sensors
ices to your friends based upon the access control policies and actuators, and the services that this enables.
you set. For the Web of Things, personal zones offers a uni-
fied means for people to manage the data they own. Links:
IERC 2104: European Research Cluster on the Internet of
Another consideration is the lifecycle of services, their provi- Things, 2014: Internet of Things From Research and
sioning, and packaging as products for sale to consumers. As Innovation to Market Deployment:
an example, consider a security camera purchased for use at http://www.internet-of-things-research.eu/pdf/IoT-
home. The camera may be bundled with the software and From%20Research%20and%20Innovation%20to%20Mark
services, or this could be purchased separately from a Web of et%20Deployment_IERC_Cluster_eBook_978-87-93102-
Things marketplace. Either way, the user needs a really 95-8_P.pdf
simple approach to installing the hardware and setting up the
associated services. How is the camera ‘discovered’ and COMPOSE: EU FP7 project 2012-2015,
‘paired’ with a service? How does the user provide additional http://www.compose-project.eu
metadata, e.g., giving it a name, describing its location, and
setting its access control policy? The package could include Webinos: EU FP7 project 2010-2013,
an application that the user runs to set this up, and to install http://webinos.org
any required device drivers. The package could include other
applications that enable the user to manage the device and Cisco, 2011: The Internet of Things — How the Next
service, and as a user interface for the service when it is up Evolution of the Internet Is Changing Everything:
and running. In the context of a smart city, there may be a http://www.cisco.com/web/about/ac79/docs/innov/IoT_IBS
need to install and set up large numbers of sensors. This too G_0411FINAL.pdf
should be as simple and painless as possible. The same
should be true for managing software upgrades and for taking IoT2014: David Clark — Not making the same mistakes
devices and services out of service as needed. again:
http://www.iot-conference.org/iot2014/keynote-speakers/
Suppliers and consumers of services need to reach agree-
ments, and this can be formalised as contracts that cover pay- Google, 2014: The Physical Web
ments, data handling policies and so forth. For open markets https://google.github.io/physical-web/
of services such contracts should be legally binding on the
parties involved. Whilst data may be provided free, in other
cases, some form of payment will be required, for instance, Please contact:
one off payments, per usage payments and subscription based Emmanuel Baccelli, Inria, France
payments. To enable open markets to operate across borders, E-mail: Emmanuel.Baccelli@inria.fr
there is a need for international standards around payments.
Even if services are provided free of charge, they may be Dave Raggett, W3C
restricted to consumers in certain groups. Access control may E-mail: dsr@w3.org
be based on rules, e.g., as with the XACML XML access con-
trol language, or based upon the possession of tokens as with
capability based systems.
Time Synchronized Channel Hopping networking solutions for Low-Power allows TSCH to fit under an IPv6-
(TSCH) was designed to allow and Lossy Networks (LLNs). The enabled protocol stack for LLNs, run-
IEEE802.15.4 devices (i.e., low power memo RFC5673 (Industrial Routing ning IPv6 packet delivery in Low
wireless mesh network devices) to sup- Requirements in Low-Power and Lossy Power Wireless Personal Area
port a wide range of applications Networks) (Link 2) discusses industrial Networks (6LoWPANs) (Link 3), IPv6
including, but not limited to, industrial applications, and highlights the harsh Routing Protocol for LLN (RPL) (Link
automation and process monitoring. It is operating conditions as well as the strin- 4) and the Constrained Application
based on a medium access technique gent reliability, availability, and security Protocol (CoAP) (Link 5). All of this
which uses time synchronization to requirements for an LLN to operate in complemented with the management
achieve ultra low-power operation and an industrial environment. In these sce- interface and network operation specifi-
channel hopping to enable high relia- narios, vast deployment areas with large cation currently being developed by the
bility. Synchronization accuracy directly (metallic) equipment cause multi-path IETF 6TiSCH working group.
relates to power consumption, and can fading and interference to thwart any
vary from microseconds to milliseconds, attempt of a single-channel solution to The OpenWSN project (Link 6) (see
depending on the hardware and imple- be reliable; the channel agility of TSCH Figure 1) is an open-source implemen-
mentation. is the key to its ultra high reliability. tation of the aforementioned protocol
stack designed for capillary networks,
The IEEE802.15.4e standard is the latest IEEE802.15.4e TSCH focuses on the rooted in the new IEEE802.15.4e TSCH
generation of ultra-low power reliable MAC layer only. This clean layering standard, and providing IPv6 connec-
tivity to ultra-reliable low power indus-
trial mesh networks [1].
Figure 1: OpenMote
CC2538 main OpenWSN was founded in 2010 in the
board. Composed by Berkeley Sensor & Actuator Center at
a ARM Cortex M3 UC Berkeley as an open-source project
SoC TI CC2538 to implement, promote and contribute
Microcontroller. to the Industrial Internet of Things
Open-Hardware (IIoT), the next wave of innovation
design. impacting the way the world connects.
With a community of academic and
industrial contributors, the OpenWSN
protocol stack now runs on most pop-
ular low-power wireless platforms, and
features a simulator and an ecosystem
of cloud-based collaboration and con-
tinuous integration tools. OpenWSN
has become the de-facto open-source
implementation of IEEE802.15.4e-
2012 Time Synchronized Channel
Hopping (TSCH), the standard at the
heart of the IIoT which enables ultra
high reliability and low-power opera-
tion.
6TiSCH is spearheading the develop- viral nor restrictive, enabling industrial 6. OpenWSN Project:
ment of the Industrial IoT. users to take advantage of the platform http://openwsn.org/
without jeopardizing their develop- 7. http://tools.ietf.org/wg/6tisch/charters
In 2014, members of the OpenWSN ments or intellectual property. 8. OpenMote Technologies
team founded OpenMote (Link 8), a OpenWSN is in constant evolution, http://www.openmote.com/
startup company developing an being updated with the latest standards
ecosystem of hardware for the IoT, in the field and becoming a central pro- Reference:
including the popular OpenMote (see totyping platform for future amend- [1] T. Watteyne et al.: “OpenWSN: A
Figure 2), the OpenBase/OpenUSB ments and improvements to already Standards-Based Low-Power Wireless
board to program it, and the existing standards and protocols. Development Environment”,
OpenBattery to power it. Inheriting from Transactions on Emerging
previous ‘Berkeley Motes’, the Links: Telecommunications Technologies,
OpenMote is widely considered the new 1. http://www.openmote.com/ vol. 23, issue 5, 480-493, 2012.
generation of low-power wireless exper- 2. Industrial Routing Requirements in
imental platforms. Low-Power and Lossy Networks: Please contact:
https://tools.ietf.org/html/rfc5673 Thomas Watteyne
The OpenWSN ecosystem is devoted to 3. Compression Format for IPv6 Inria, HiPERCOM2 team
accelerating the standardization and Datagrams over IEEE 802.15.4-Based E-mail: thomas.watteyne@inria.fr
transfer of low power wireless technolo- Networks:
gies to the industry, enabling and pro- https://tools.ietf.org/html/rfc6282 Xavier Vilajosana
moting the adoption of low power wire- 4. RPL: IPv6 Routing Protocol for Universitat Oberta de Catalunya
less for industrial applications. Low-Power and Lossy Networks: E-mail: xvilajosana@uoc.edu
OpenWSN is licensed under the BSD https://tools.ietf.org/html/rfc6550
license; OpenMote Hardware is 5. The Constrained Application Pere Tuset
licensed under CERN OHL 1.2 hard- Protocol (CoAP): OpenMote Technologies
ware license. Both licences are neither https://tools.ietf.org/html/rfc7252 E-mail: peretuset@openmote.com
The Internet of Things (IoT) is expected to be the next ‘big thing’. To date, however, there is no de
facto standard software platform to program memory and energy-constrained IoT devices [1]. We
expect an evolution of IoT software platforms that can be roughly compared to the recent evolution
of software platforms powering smartphones.
Over the course of a few years, there has installed/upgraded billions of times on lacking enough openness. In contrast,
been an acceleration in the evolution of smartphones and tablets. Arguably, the prominent examples have achieved
software platforms powering smart attribute of (re)programmability has high impact in the domain of computer
handheld devices - from a state where had the most significant impact in this networks, while remaining open, and
dozens of closed-source, slow-progress, field, even more so than the evolution of therefore more trustworthy: for
rudimentary proprietary solutions where handheld hardware. instance, community-driven efforts
used, to a state where just two major such as Linux, or the IETF [2], each
players (iOS and Android) have imposed On the dark side, companies pushing fundamental to the emergence of the
new de facto standards in terms of soft- iOS and Android use such software Internet as we know it today. While
ware platform interoperability, program- platforms as tools to achieve question- Linux demonstrated the power of open
mability and automatic updating. able goals e.g., more or less covertly source software development and
accumulating and monetizing personal- became the go-to operating system for
The up side of this evolution is quicker- ized information. Furthermore, the true Internet hosts, the IETF is an out-
paced progress and the development of level of openness of iOS or Android is standing example of the positive impact
innumerable applications built on top of debatable, and their not being fully open of open standards, transparent standard-
these platforms, which interact in inno- facilitates the task of covertly accumu- ization processes, and open discussion
vative ways with one another, and with lating personal information. The public forums.
the cloud. It has become so easy to is increasingly aware of this pitfall in
develop such applications (the infamous terms of security and privacy, and it is Until recently, IoT software platforms
‘Apps’) that hundreds of thousands of therefore unlikely that people will were in a comparable state to that of
developers have been able to produce accept IoT software platforms intrinsi- smartphone software platforms before
millions of Apps, which have been cally tied to such monetization, or Android and iOS. It is likely, however,
that new standards will be set in the near with no learning curve (assuming prior cations are expected to leverage sponta-
future, in terms of IoT software plat- experience with POSIX and Linux). On neous wireless networks, Internet con-
form openness, API, automated soft- a wide variety of IoT hardware, RIOT nectivity, the cloud and a dense, inter-
ware updates and other characteristics enables the use of the standard coding connected environment of communi-
that are necessary to enable a modern, languages (ANSI C and C++), well- cating devices. In the mid and long
large scale, secure ecosystem. Such an known debugging tools (gdb, Valgrind term, such IoT applications are
evolution will be game-changing for the etc.), and standard programming para- expected to not only enable optimiza-
IoT, and will fuel a new world of dis- digms (such as full multi-threading), tion of existing processes, but also
tributed applications developed by a while being energy efficient and real- entirely new processes and the emer-
large variety of actors, on top of the time capable. One way to gauge this gence of an enhanced reality, in which
dominant open software platform(s). As effort is to make a parallel with IETF our interface with the Internet will no
yet it is unclear which IoT software protocols such as 6LoWPAN or CoAP, longer be predominantly a screen, but
platforms will emerge as dominant. which adapt IP to memory and energy- rather the objects of the cyber-physical
constrained IoT devices, without losing system embodied by the Internet of
In practice, IoT software platforms face interoperability with the rest of the Things itself.
conflicting requirements: interoper- Internet. Similarly, a modern IoT oper-
ability with the Internet, memory-con- ating system such as RIOT enables a Link:
strained embedded programming, and full-fledged operating system on IoT The SAFEST Project, co-funded by
portable, open-source code. Over the devices with resources that are too lim- ANR and BMBF,
last decade, significant progress has ited for Linux to run on, without losing http://safest.realmv6.org
been made in order to accommodate interoperability with state-of-the-art
such requirements. Pioneer open source development tools and programmers. References:
software platforms such as Contiki or [1] C. Bormann et al.: “Terminology
TinyOS provided first attempts at RIOT is developed by an international for Constrained node Networks", RFC
accommodating these requirements, by community of open source developers 7228, Internet Engineering Task Force,
exploiting non-standard coding para- (co-founded in 2013 by Inria, Freie 2014.
digms, or by limiting code portability Universität Berlin, and Hamburg [2] The Internet Engineering Task
and the set of functionalities offered by University of Applied Sciences in the Force (IETF), http://www.ietf.org
the software platform (for instance, context of the SAFEST project), using [3] RIOT: The Friendly Operating
Arduino environment). Recently, more transparent decision processes based on System for the Internet of Things.
powerful but comparably low memory- rough consensus, open discussion http://www.riot-os.org
footprint software platforms became forums, open-source code, and open
available. standards. The near-future goal of RIOT Please contact:
is to power a modern, large scale, evolu- Emmanuel Baccelli
A prominent example is RIOT [3], an tionary, and secure cyber-physical Inria, France.
open source IoT operating system ecosystem, comprising heterogeneous E-mail: emmanuel.baccelli@inria.fr
which enables programmers to develop IoT devices, distributed processes and
applications on typical IoT devices, applications (see Figure 1). Such appli-
FIT IoT-LAB:
The Largest IoT Open Experimental Testbed
by Eric Fleury, Nathalie Mitton, Thomas Noël and Cedric Adjih
The universal proliferation of intelligent objects is making Internet of Things (IoT) a reality; to operate on a
large scale it will critically rely on new, seamless, forms of communications. But how can innovations be
validated in a controlled environment, before being massively deployed into the real world? FIT IoT-LAB
addresses this challenge by offering a unique open first class service to all IoT developers, researchers,
integrators and developers: a large-scale experimental testbed allowing design, development, deployment
and testing of innovative IoT applications, in order to test the future and make it safe.
IoT is not simply emerging as a major development, tuning, and experimenta- respect to different features, such as
technology trend: it is already a reality tion related to IoT. microcontroller (TI MSP430, ARM
with billions of existing devices. The Cortex M3 and ARM Cortex A8), radio
Internet of Things represents a tremen- IoT-LAB is part of the FIT (Future chip (2.4GHz or 800MHz) and addi-
dous paradigm shift since Internet was Internet of Things) project which tional functionalities (mobility / robots,
designed; an evolution from pure end- develops experimental facilities within a accelerometer, magnetometer and
to-end communication between an end- federated and competitive infrastructure gyrometer). Resources can be reserved
user device and a server in the Internet, with international visibility and a broad on one or several sites at once. Services
to an Internet interconnecting physical panel of customers. All facilities come offered by IoT-LAB include:
objects that are freely able to communi- with complementary components that • Total remote access to nodes
cate with each other and with humans. enable experimentation on innovative reserved, e.g., allowing users to flash
IoT builds on three pillars [1], related to services for academic and industrial any firmware, without any condition
the ability of smart objects to: (i) com- users. FIT has received 5.8 million Euros or constraint. Any language or OS
pute, (ii) communicate, and (iii) sense in funding from the Equipex research can be used to design, build, and
and interact with their environment. grant program. The FIT consortium is compile applications;
coordinated by University Pierre et • Direct access to a debugger server on
Although IoT is already a reality, it is Marie Curie and composed of Inria, each node so that all debugging can
still maturing and waiting for its ‘iPhone ICube laboratory from University of be performed remotely on the node
moment’. Several challenges remain, in Strasbourg, Institut Mines-Télécom and (such as step by step code execution)
particular relating to the standardization CNRS. FIT is a platform federation. • Access to the serial ports of all nodes
of efficient and universal protocols, and Such a federation of independent net- for a real-time interaction, with
to the design and testing of IoT services work experimental facilities is arguably optional aggregation;
and applications. Owing to their mas- the only meaningful way to achieve the • Each node could be visible from
sively distributed nature, their design, required scale and representativeness for Internet with end-to-end IP connec-
implementation, and evaluation are supporting Future Internet research. IoT- tion using IPv6 and 6LoWPAN for
inherently complex and tend to be LAB testbeds are dispersed among six example;
daunting, time-consuming tasks. different locations across France offering • A very accurate power consumption
Required to overcome this hurdle is a access, for the first time, to 2728 wireless monitoring of every node.
representative, large scale, platform IoT fixed and mobile nodes equipped • Packet sniffer and analyzer on each
allowing researchers, IoT designers, with various sensors (see Table 1). node;
developers and engineers to construct, • A GPS module for some A8 nodes
benchmark and optimize their protocols, Users can select and reserve the number allowing a very precise end-to-end
applications and services. and type of nodes they wish with time synchronization, accurate moni-
toring and performance evaluation of plug their own hardware devices, Reference:
communication protocols; while benefiting from the IoT-LAB [1] D. Miorandi, S. Sicari, F. D.
• A set of useful detailed tutorials, OS services and monitoring tools (a fea- Pellegrini, I. Chlamtac: “Internet of
supports (Contiki, FreeRTOS, TinyOS, ture frequently required by both aca- things: Vision, applications and
and RIOT) including full protocol demic and industrial users). research challenges”, Ad Hoc
stacks and communication libraries Networks, 10(7):1497 – 1516, 2012.
such as OpenWSN providing open- Overall, FIT-IoT LAB testbed is a
source implementations of IoT proto- unique pioneer in the domain of IoT test- Please contact:
col standards; beds. Eric Fleury
• A unique fleet of mobile robots ENS de Lyon / Inria, France
(WifiBot and TurtleBot); Links: Tel: +33 672 162 974
• Strong extensibility through the IoT-LAB: https://www.iot-lab.info E-mail: Eric.Fleury@inria.fr
availability of more than 100 empty FIT: https://www.fit-equipex.fr
slots on which users can physically OpenWSN: https://openwsn.atlassian.net/
OCARI is a wireless sensor network designed to operate in industrial environments [1]. It is easy to
deploy (i.e. ‘plug-and-play’), and is energy-efficient to support battery-operated nodes. OCARI nodes
use commercial off-the shelf components. OCARI provides bounded medium access delays and the
energy consumption of an OCARI network is predictable. In addition, the network is scalable (up to
hundreds of sensor nodes) and able to support micro-mobility of nodes.
OCARI can be distinguished from Furthermore, OCARI has the advantage includes OPERA (OPtimized Energy
WirelessHart and ZigBee by the fol- of being open source. efficient Routing and node Activity
lowing three characteristics: scheduling) that consists of:
• It relies on a mesh topology (see Fig- The OCARI stack comprises the • EOLSR, an energy-efficient routing
ure 1), improving robustness and Physical, MAC, Network and including neighbourhood discovery
bandwidth use. Application layers. It was first devel- and building of the routing tree root-
• It is self-configuring thanks to a oped in the ANR OCARI project lead ed at the sink in charge of data gath-
dynamic multihop routing taking by EDF with Inria, LIMOS, LATTIS, ering.
energy into account. LRI, Telit and DCNS. Today, it is • OSERENA, [2] a colouring node pro-
• It saves energy thanks to a determinis- actively developed by EDF, Inria, in tocol allowing a conflict-free sched-
tic medium access and sleeping peri- collaboration with ADWAVE and other ule of node medium accesses to be
ods for nodes. research labs. The network layer built. Each node knows the time slots
in which it is allowed to transmit or it the 2.4GHz band and a 32 bits Cortex • Demonstration during the Inria-
may receive from a neighbour. In the M3 microcontroller. OCARI is presently Industry Telecom day: routing and
remaining slots it sleeps to save energy. available on two platforms: Dresden colouring, fire detection on a DCNS
deRFsam3-23T09-3 and ADWAVE ship construction site using
OCARI takes advantage of cross-lay- ADWRF24-LRS (see Figure 2). ADWAVE hardware (see
ering [3] to optimize its performance. http://www.inria.fr/en/innovation/
From the hardware point of view, an OCARI has been demonstrated many calendar/telecoms-du-futur).
OCARI node consists of a radio fre- times, the most recent two demonstra- • Demonstration for the steering com-
quency transceiver compliant with the tions in November 2014: mittee of the Connexion Cluster proj-
IEEEE 802.15.4 standard operating in ect (https://www.cluster-connexion.fr):
integration of sensors of various
types in the OCARI network and
interconnection of the OCARI net-
work to the facility backbone by
Figure 1: OCARI topology. means of a OCARI/OPC-UA gate-
way built with Telecom ParisTech on
a Raspberry Pi (see Figure 3).
Link: http://www.ocari.org
Please contact:
Tuan Dang
EDF, STEP department, France
E-mail: tuan.dang@edf.fr
The Internet of Things (IoT) has particular security and privacy problems. The Internet Engineering
Task Force is designing authentication and authorization mechanisms for the most constrained
devices which are part of the Internet of Things.
There exists a multitude of implemented, as well as envisioned, use cases for the Internet of Things
(IoT) and Wireless Sensor Networks (WSN). Some of these use cases would benefit from the
collected data being globally accessible to: (a) authorized users only; and (b) data processing units
through the Internet. Much of the data collected, such as location or personal identifiers, are of a
highly sensitive nature. Even seemingly innocuous data (e.g., energy consumption) can lead to
potential infringements of user privacy.
The infrastructure of the Internet of resources of these devices, the goal of SecureWSN tackles this challenge by
Things (IoT) with its diversity of devices secure WSNs is to support end-to-end developing three solutions for different
and applications, as well as the trend security by a two-way authentication, types of resources.
towards a separation of sensor network an efficient data transport solution for
infrastructure and applications exacer- the data, and a controlled data access, TinyDTLS protects data from its source
bates security risks. A true end-to-end supporting the mobility of today’s to the sink supporting confidentiality,
security solution is therefore required to users. Thus, different components were integrity, and authenticity. RSA-capable
achieve an adequate level of security for developed in the construction of (Rivest-Shamir-Adleman) devices are
IoT. Protecting data once they leave the SecureWSN and are illustrated in authenticated via X.509 certificates
boundaries of a local network is not suf- Figure 1. All components support hard- during the key-exchange in a two-way
ficient, especially when private and high- ware from different vendors with dif- authentication handshake. Constrained
risk information are effected. ferent resources. Different security devices perform a variant of the
solutions (TinySAM, TinyDTLS [2], or Transport Layer Security (TLS) pre-
However, IoT is no longer limited to TinyTO [3]) were developed that are shared key algorithm. The data sink
servers, routers, and computers with based on known algorithms from IP net- authenticates via certificate either
manifold resources. It also includes con- works, like DTLS and BCK, and directly with the mote or with an Access
strained devices – motes –, which are required adaptation (e.g., complexity) Control Server (ACS). ACS grants
very limited in memory (approximately to fit these resources whilst supporting a tickets to authenticated devices with suf-
10-50 KByte RAM and 100-256 KByte heterogonous network structure. ficient rights. Motes request connection
ROM), computational capacity, and from their communication partner where
power (supported by just a few AA bat- End-to-end Security key establishment is based on DTLS [2].
teries). These limited resources do not Today, with so much personal informa-
reduce the need to support end-to-end tion online, end-to-end security is In comparison, TinyTO uses a Bellare-
security and secure communications, but essential in many situations. This repre- Canetti-Krawcyk (BCK) handshake
they make it much harder to meet these sents the challenge for constrained with pre-shared keys. For the key gener-
requirements. Depending on the specific devices usually used in WSNs. ation, key exchange, signatures, and
Reactive Blocks is a toolkit to create, analyse and implement reactive applications. It helps
developers to get concurrency issues right, reuse existing solutions, and brings model checking
to the common programmer.
Internet of Things (IoT) applications Model-driven development and model often fail to cover details in code in a
connect hardware to communicate checking provide solutions to these suitable way.
within a network, often with constrained problems. However, these solutions are
resources. Therefore, even simple use barely used by programmers. Reasons With these concerns in mind, we devel-
cases quickly become very complex. In for this are that model checking oped the Reactive Blocks tool for event-
addition, IoT applications often combine requires deep knowledge in formal driven and concurrent applications. It
several technologies, and only few pro- methods to produce an appropriate has its origins at the Department of
grammers have the skill set to cover input model that can be analysed. In Telematics at the Norwegian University
them all. addition, model-driven approaches of Science and Technology (NTNU),
1 start
start
Config
Config Listener
Listener 2 3
initConfig
initConfig
start updatedConfig
updatedConfig
init
CloudClient
Kura Clo udClient
Modbus
Modbus
publish
publish ready
d
executedReadIn
executedReadIn iinitOk
nitOk
deviceId
deviceId rreadDone
eadDone waiting
waiting
failed
e xecutedWrite
executedWrite
readFailed
stop
stop writeDone
writeDone stop
stop stopped
e xecuteReadHold
executeReadHold
playing msgArrived init
rregisterId
egisterId
writeFailed
writeFailed CoAP Resource
d
stopped
start ok
initFailed
init
Config
Config Listener
Listen
MQTT Core get
start
Config Listener
ener
subscribe
sub scribe iinitOk
nitOk initConfig
initCo
initConfig
Config
post
publishOk
publishOk updatedConfig
Config
unsubscribe
unsub scribe
updatedConfig
updatedCo getResponse
error
publish
publish failed
message
me ssage init initPS
RPi G
RPi PIO Di
GPIO Digital n
gital IIn
stop
stop disconnected
d isconnected
stop
stop value
value
isLow
initError stopped
Figure 1: The Reactive Blocks workflow. Blocks from libraries are combined to applications. Blocks are model-checked, and the application built
via code generation.
and has been further developed by the is usually the kind of code that is cum- system is informed about the arrival of
spin-off company Bitreactive. bersome and error-prone to write manu- freight. In case of deviations, alerts via
ally. email and SMS are sent. The system
Building IoT Applications with was created by programmers without
Reactive Blocks Compositional Analysis specific knowledge in formal methods.
To appeal to the common programmer, A model checker is built into the editor. Due to the built-in analysis, all interac-
the tool combines programming and The building blocks have a formal tions work correctly. Using building
modelling with an automatic composi- semantics and correspond to temporal blocks, about 60% of the functionality
tional analysis, as shown in Figure 1: A logic formulas, so that the model could be taken from reusable libraries.
developer creates a system from checker can take the UML model as
building blocks. Many of the blocks do input. Developers can start model Link:
not need to be developed from scratch checking with a single click. The encap- http://bitreactive.com
but can be selected from libraries (1) sulation of building blocks by contracts
The blocks are connected with each enables a compositional analysis, in References:
other (2) This is facilitated by behav- which each block can be analysed sepa- [1] L.A. Gunawan, P. Herrmann:
ioural contracts enabling the developer rately. This minimises the problem of ”Compositional Verification of
to understand the interface behaviour of state space explosion. An error is shown Application-Level Security Properties,
a block without having to understand its to the developer as animation in the in proc. of ESSoS 2013, LNCS 7781,
inner details. Applications consist of a editor such that can be easily under- February/March 2013.
hierarchy of blocks. The applications stood and fixed. Examples of errors are
are then analysed by automatic model deadlocks or violations of protocols, [2] Han F., P. Herrmann: ”Modeling
checking. Once consistent, code is auto- which are detected because the UML Real-Time System Performance with
matically generated and can be model of the building blocks do not Respect to Scheduling Analysis”, in
deployed as OSGi bundles or stand- reach their final state, or because the proc. of UMEDIA 2013, Aizu-
alone Java applications (3). external contract of a block is violated. Wakamatsu, IEEE Computer Society
In addition, the tool detects if an appli- Press, September 2013.
To balance the benefits of generating cation does not honour the life cycle
code with that of manual programming, contract of a component when it runs in [3] P. Herrmann, J.O. Blech, F. Han, H.
Reactive Blocks uses a combination of a framework like OSGi. Such errors are Schmidt. A Model-based Toolchain to
Unified Modelling Language (UML) hard, if not impossible, to detect by Verify Spatial Behavior of Cyber-
with Java. UML describes the concur- testing. The mathematical basis of the Physical Systems. In 2014 Asia-Pacific
rent behaviour of building blocks with building blocks in temporal logic for- Services Computing Conference
activity diagrams or state machines. mulas makes it possible to reason about (APSCC), IEEE Computer.
State machines are also used to describe other properties, such as reliability,
the contract of a block. The UML security [1], real-time properties [2], or Please contact:
models only coordinate concurrent spatial constraints [3]. Frank Alexander Kraemer, Peter
behaviour, and refer to Java methods for Herrmann
detailed operations on application pro- Fleet Monitoring Case Study NTNU, Norway,
gramming interfaces (APIs) or other A case study was developed within the E-mail: kraemer@item.ntnu.no,
data. The Java methods are edited in the IMSIS project funded by the Norwegian herrmann@item.ntnu.no
Java tools within Eclipse. Existing code Research Council. It consists of a
can also be integrated, by encapsulating system that monitors a fleet of trucks
it into operations and building blocks. while they deliver goods. Their position
The code related to concurrency, i.e., relative to defined geo-fences is con-
code that decides when the operations stantly monitored and compared to
are called, is generated from UML. This expected arrival times. The backend
XForms is a language for describing interfaces to data, designed at W3C by researchers from
industry and academia. It is a declarative language, meaning it describes what has to be done, but
largely not how. The interface it describes does not have to run locally on the machine producing the
data, but can be run remotely over the network. Since Internet of Things (IoT) computers typically
have little memory and are low-powered, this makes XForms ideally suited for the task.
One of the unexpected successes of XForms has already been used for a administrative programming needed.
HTML was its adoption for controlling number of years to control devices in Several implementations of XForms are
devices with embedded computers, such this way at many petrol stations in the available, at least three of which are
as home Wi-Fi routers. To make an USA. Each device, storage tanks, petrol open-source. It is also part of the Open
adjustment to such a device, the user pumps, cash registers, and so on, con- Document Format, implemented by
directs the browser to the IP address tains a simple server that delivers its Open Office and Libre Office. There is
from which it is running and a small web data as XML instances. XForms inter- also a tutorial [3].
server on the device serves up web faces are then used to read and combine
pages that allow the user to fill in and these values, and update control values Link:
submit values to change the working of (for instance the price of fuel being dis- Nest API Reference:
the device. played on pumps). https://developer.nest.com/documentation/
api-reference
However, the tiny embedded computers As an example of how it could be used,
that form part of the IoT typically have Nest, a well-known producer of internet References:
memory in kilobytes, not megabytes, thermostats, has published the data- [1] J. M Boyer (ed.), XForms 1.1,
and lack the power to run a web server model interface to its devices. A simple W3C, 2009,
that can serve and interpret web pages. interface to this could look like this: http://www.w3.org/TR/2009/REC-
This calls for a different approach. xforms-20091020/
<instance src=”http://thermostat.local/”/> [2] J. M. Boyer et al. (eds.), XForms
One approach is for the devices to serve <bind ref="ambient_temperature_c" 2.0, W3C, 2015,
up only the data of the parameters, so type="decimal" readonly="true()"/> http://www.w3.org/MarkUp/Forms/wiki/
that those values can then be injected <bind ref="target_temperature_c" XForms_2.0
into an interface served from elsewhere. type="decimal"/> [3] Steven Pemberton, XForms for
XForms [1], a standard that we have <bind ref="target_temperature_f" HTML Authors, W3C, 2010,
helped develop at W3C, is designed for type="decimal" calculate= http://www.w3.org/MarkUp/Forms/2010/
exactly this type of scenario: although it "../target_temperature_c*9/5+32"/> xforms11-for-html-authors
is a technology originally designed for <submission resource=
improving the handling of forms on the "http://thermostat.local/data" Please contact:
web, it has since been generalised to method="put" replace="instance"/> Steven Pemberton
more general applications; version 2.0 is Chair of the W3C XForms WG, CWI,
currently in preparation [2]. The following device-independent user The Netherlands
interface control specifies that a single E-mail: steven.pemberton@cwi.nl
XForms has two essential parts: the first value is to be selected from the list of
part is the model that specifies details of items, without specifying how that is to
the data being collected, where it comes be achieved (using radio buttons, drop-
from, its structure, and constraints. It downs or whatever):
allows data from several sources to be
combined, and data to be submitted to <select1 ref="temperature_scale"
different places. It also ensures that as label="Scale">
data is changed, that relationships <item value="f" label="°F"/>
between the data are kept up to date. The <item value="c" label="°C"/>
second part is the user interface that dis- </select1>
plays values and specifies controls for
changing the values. These controls are Experience with XForms has shown
specified in a device-independent that using it can reduce production time
manner that only describes what they are and costs by a factor of ten. As an
meant to achieve, not how to do it. This example, one very large pilot project
makes it easier to adapt the interface to reduced production time from five years
different devices, screen sizes, etc., with thirty people to a single year with
while still allowing the use of specific ten people. These remarkable savings
interfaces, such as radio buttons, via are due largely to the declarative nature
style sheets. of XForms, which greatly reduces the
We present a practical and scalable solution that aims to achieve the Internet of Things (IoT) paradigm in
complex contexts, such as the home automation market, in which problems are caused by the presence of
proprietary and closed systems with no compatibility with Internet protocols.
References:
[1] V. Miori, D. Russo, M. Aliberti:
“Domotic technologies incompatibility
becomes user transparent”,
Communications of the ACM, vol. 53
(1) pp. 153 - 157. ACM, 2010.
[2] M. Jung et al.: “Heterogeneous
device interaction using an IPv6
enabled service-oriented architecture
for building automation systems”, in
proc. of SAC’13, ACM, 1939-1941,
2013, DOI=10.1145/2480362.2480722
delivered to the TechManager and, after A future development will be to decen- [3] G. Briscoe, P. De Wilde: “Digital
the DomoML translation process, is for- tralize the entire DomoNet architecture ecosystems: evolving service-
warded to the DomoNet server, which in order to create virtual independent orientated architectures”, in proc. of
then sends it to the requester. devices that act as agents capable of BIONETICS’06, ACM, , Article 17,
cooperating with each other according 2006. DOI=10.1145/1315843.1315864
We are working on the development of to the Digital Ecosystem architecture http://doi.acm.org/10.1145/1315843.13
a web server running inside the same paradigm [3]. This approach would fur- 15864
computer that hosts the DomoNet ther the realization of the concepts
server, in order to simulate a direct web underlying the IoT paradigm and pro- Please contact:
interaction with the device identified by mote a new model for thinking about Vittorio Miori
its IPv6 address. The Tomcat web server the environment surrounding us and the Domotics Lab, ISTI CNR, Italy
was used in the test architecture. objects contained therein. E-mail: vittorio.miori@isti.cnr.it
The Dioptase middleware provides developers with new methods for writing distributed
applications for the Internet of Things (IoT). Dioptase leverages data streams as data model and
continuous in-network processing as computation model, in order to deal with the challenging
volume of data being continuously produced by the ‘Things’.
The Internet of Things (IoT) is currently devices, mobile versus fixed devices, lems emerge from these approaches,
characterized by an ever-growing continuously-powered versus battery- related to energy consumption (due to
number of networked ‘Things’, i.e., powered devices, etc.). These chal- intense wireless communications) and
devices that have their own identity lenges require new systems and tech- device lifetime, network overload and
together with advanced computation and niques for developing applications that privacy preservation. The challenge is
networking capabilities: smartphones, are able to: (i) collect and process data then to allow Things to be much more
smart watches, smart home appliances, from the numerous data sources of the autonomous and use third-party serv-
etc. These Things are being equipped IoT and (ii) interact both with the envi- ices only when required, such as in
with increasing numbers of sensors and ronment using the actuators, and with wireless sensor networks, but at a much
actuators that enable them to sense and the users using dedicated GUIs. larger scale.
act on their environment, linking the
physical and virtual worlds. Specifically, Solutions for the IoT currently rely To this end, the Dioptase [1] solution
the IoT raises many challenges related to heavily on third-party services and aims at providing a common middle-
its very large scale and high dynamicity, infrastructures, such as the cloud of ware layer that runs directly on the
as well as the great heterogeneity of the Things where each device sends its Things, enabling them to manage the
data and systems involved (e.g., pow- measurements to centralized collection huge volume of data continuously being
erful versus resource-constrained and computation points. Several prob- produced (measurements, events, logs,
etc.) in a collaborative way. In this con- expressed by the developers and the gate, adapt and split their own tasks
text, we consider data streams and con- characteristics of the available Things. according to their environment, their
tinuous processing as the reference data load, their available resources and their
and computation models for developing Dioptase features a customizable mid- capabilities in a decentralized collabo-
IoT applications. Continuous pro- dleware architecture that is versatile rative manner, taking the role of DiDS.
cessing is indeed a very suitable para- enough to be deployed on a large class As a benefit, the network of Things
digm for processing each piece of data of Things that vary significantly in would adapt itself to minor and major
one by one, without having to store the terms of resource availability (e.g., changes over time.
entire dataset.
• processors, which produce new data Runs directly onto Things and abstracts them
Developers
Developers Deployment and
streams by continuously processing or as generic pool of sensing, actuation, orchestration servers
or
Users
computing and storage resources
existing ones; Users
Once Dioptase is deployed onto a Figure 1: Dioptase, a data streaming middleware for the IoT.
Thing, it enables developers to manage
the Thing as an abstracted and homoge-
neous pool of resources that can execute embedded systems, smartphones or
stream services provided over time. plug computers), provided that these
These stream services can be developed Things are able to communicate directly
by using the native languages of the through the Internet infrastructure. As
device or a new lightweight stream pro- illustrated in Figure 1, Dioptase embeds
cessing language, called DiSPL, various modules for interacting with
directly interpretable by Dioptase. existing components being used as part Link:
of the IoT infrastructure: discovery sys- https://mimove.inria.fr/dioptase
By composing these stream services, tems and registries, computation and
developers build their IoT applications offloading infrastructure, legacy sensor References:
as a set of tasks that are executed con- and actuator networks, etc. [1] B. Billet, V. Issarny: “Dioptase: A
tinuously by the hosting Things in a distributed data streaming middleware
purely distributed manner. Developers This work was part of the larger for the future Web of Things”, Journal
are provided with dedicated tools for CHOReOS project effort, which revisits of Internet Services and Applications,
designing tasks graphically before the concept of choreography-centric vol. 5, no. 13, 2014.
injecting them into the network of service-oriented systems to introduce a [2] B. Billet, V. Issarny: “From task
Things at any time, using deployment dynamic development process and graphs to concrete actions: A new task
Web services or through the Dioptase associated methods, tools, and middle- mapping algorithm for the future
deployment server (DiDS). DiDS auto- ware for the services in the Ultra Large Internet of Things”, in proc. of the
matically computes where to deploy the Scale (ULS) Future Internet. Beyond 11th IEEE International Conference on
stream services and then manages their CHOReOS, Dioptase represents a sig- Mobile Ad hoc and Sensor Systems
execution over time. DiDS relies on a nificant step towards an IoT where (MASS), 2014.
state-of-the-art task allocation algo- every Thing can be assigned tasks and
rithm [2], which guarantees the lifetime complete them autonomously. As an Please contact:
of each task, by minimizing the overall ongoing area of research, Dioptase will Benjamin Billet, Valérie Issarny
energy consumption or maximizing the be released soon as an open source Inria, France
fair allocation of tasks among the project. For the near future, we plan to E-mail: Benjamin.Billet@inria.fr,
Things, according to the constraints enable Things to automatically dele- valerie.issarny@inria.fr
The Internet of Things (IoT) enables a large number of devices to cooperate to achieve a common
task. Each individual device is small and executes a confined software core. Collective intelligence is
gained from distributed collaboration and Internet communication. Corresponding IoT solutions form
large distributed software systems that pose professional requirements: scalability, reliability,
security, portability and maintainability. The C++ Actor Framework CAF [1] contributes such a
professional open source software layer for the IoT. Based on the actor model of Hewitt et al. [2], it
aids programmers at a good level of abstraction without sacrificing performance.
The Internet of Things (IoT) is com- from the community. CAF enables aim at avoiding intermediate state in the
posed of many nodes, often with limited asynchronous communication between network. Letting a runtime environment
capabilities. Small software components nodes by a reliable message passing handle low-level functionality gives
that communicate via Internet protocol facility. Following the 'shared nothing' developers more time to focus on the
standards typically form a highly distrib- paradigm of actors, it provides synchro- application logic.
uted work flow among machines with nization primitives as well as error han-
minimal human interaction. Traditional dling capabilities at very low overhead.
application scenarios include sensor net- CAF runs on RIOT [3], the friendly Our C++ Actor Framework has elabo-
works that monitor data such as environ- operating system for the IoT. rated this model to allow for develop-
mental or cyber-physical conditions. In ment of native software at a high
addition to sensors, IoT networks As early as 1973, Hewitt et al. [2] pro- abstraction layer without sacrificing
include actuators that influence the envi- posed the (abstract) actor model to performance. We provide exchangeable
ronment. Complex applications are built address the problems of concurrency implementations of the runtime envi-
from these nodes, for example: home and distribution. This model defines ronment to make optimized use of
automation, and tracking of health data.
These systems enable machines to
upload data to Internet servers, a task
that originally required human interac-
tion. Thus, they allow the tracking of
data everywhere and anytime.
‘Puzzle’, a mobile end user development framework, allows the end user to opportunistically create, modify
and execute applications for interacting with smart things, phone functions and web services. The
framework simplifies development of such applications through an intuitive jigsaw metaphor that allows
easy composition on touch-based devices. Users immediately identified the utility of this feature, and found
it easy to envisage using the framework in various scenarios in their daily lives.
The main purpose of this work is to the object; and more generally, to proto- seamlessly support IoT users to develop
allow users to do more with their existing type new Internet of Things (IoT) appli- applications that meet individual needs,
devices and things within their homes or cations through a user-centred design and how to create a flexible platform
at work. This can be achieved by cre- approach in which users create and which supports customisation and inter-
ating applications customized for indi- develop personalized applications. operability of IoT applications. The
vidual needs on personal mobile devices. mobile authoring environment has been
The Puzzle framework allows users to In order to allow end users to create designed taking into account how users
visualize the current status of intelligent their own IoT applications and support can foresee the functions to compose
objects, such as home appliances; how to their extensibility, the main challenges and understand how the flow of their
operate them– for instance, through a addressed in Puzzle [1] are twofold : applications progresses, adopting a
voice command; apply the commands to how to create a User Interface (UI) that metaphor that is close to users’ real life
Modern buildings are equipped with a variety of building automation technologies and sensors which are
controlled, monitored and managed through heterogeneous information systems. To date, these systems
show limited interoperability and tend to operate in isolation. The future will require a more
comprehensive and efficient operation of buildings facilitated by way of integrated and smart cooperative
management and automation systems. Web-based open-source technologies can create 3D virtual
representations of the real-time geolocated activities within a building. Such a system can simplify both
the maintenance and operation by facility managers and application services to building occupants.
The European project Building as a effective and flexible manner. Figure 1 time sensing BaaS platform. Such fea-
Service – BaaS (ITEA2 – 12011, describes the integration of value-added tures facilitate complex user-created
November 2013-October 2016) aims to visualization services on top of the ref- domain oriented applications that can
create a reference architecture and erence BaaS architecture. be used both indoors and outdoors with
service framework for future building a georeferenced environment and effi-
automation (BA) systems with extensive One of the biggest challenges for cient XML interchange [1].
integration capabilities. Based on open building automation is the collaboration
IT standards and open-source technolo- of new components and systems with This feature has been specified as a
gies, the BaaS framework enables easy existing devices and infrastructures. We functional requirement for the operation
creation of new cross-domain BA serv- contributed to this technical challenge phase of the BaaS architecture. This
ices and the largely automated integra- by creating 3D building smart objects requirement states that the architecture
tion of existing BA components in a cost with web capabilities out of the real- shall support the development of serv-
ices that use building data models
(building geometry, location and data of
rooms, floors, sensors).
Floor model data point that provides a model data of a floor within a group
es.floorModel of floors. It provides a unique identifier, a floor name, a building identifica-
tion number, a floor level, and a geometry file identification number.
Floor geometry data point that provides the geometry and additional values
es.floorGeometry of a floor. It provides a unique identifier, a floor geometry file name, a floor
geometry file size, geometry data, and the creation timestamp.
description of dynamic scenarios in vir- nance application), and interact with the
tual globes and maps. Basic architec- facility manager if the room configura- Link:
tural elements, as well as location and tion fit or not their activities and prefer- http://www.baas-itea2.eu
typology of sensors are represented in a ences (feedback application). As a
geographical context. A specific JSON result, this work enabled to increase the References:
schema was described in the Cesium usability of the smart building (increase [1] M. Jung, J. Weidinger, W. Kastner,
Language (CZML) to describe property room availability by 50%) and enhance A. Olivieri: “Building Automation and
values changing over time, such as the user experience for IoT services Smart Cities: An Integration Approach
status and information of sensors that (with a move from manual management Based on a Service-Oriented
can be provided as a stream to web to a digital system by Internet on Architecture, pp. 1361-1367, IEEE
clients to view dynamic values of building actuators). Figure 2 illustrates Computer Society (2013).
building elements. Other values trans- the 3D graphical and data visualization [2] J. Fons, D. Gaston, C. Joubert, M.
mitted from the BaaS platform to the interface part of the facility manager’s Montesinos: “Internet of Things
visualization component are the vertex maintenance application on the Patraix Applications for Neighbourhood
positions in cartographic degrees of all public building. Embedded Devices”, ERCIM News
floor rooms as well as the navigational 2014(98) (2014).
model between building, floor and We plan to deploy and test the results of [3] C. Joubert, M. Montesinos, J. Sanz:
rooms based on the location ontology. this project in the maritime sector [3] “A Comprehensive Port Operations
Among the functionalities provided to for global situational awareness in Management System”, ERCIM News
the facility manager is the dynamic link rescue, calamity and inspection opera- 2014(97) (2014)
between the status and information of tions in port infrastructures, and in the
the 3D scene with the information Smart Cities platform to scale the visu- Please contact:
panel. Hence, navigation can be done alization capabilities to more public Christophe Joubert
either through the panel’s elements, buildings. Prodevelop, Spain
such as the list of sensors, or through E-mail: cjoubert@prodevelop.es
the 3D scene billboards anywhere on In our research, we collaborated with
the building. several SMEs, Universities and
Research Centres, including: Siemens
The results to date are very promising: AG; Materna; Kieback&Peter; TWT;
we integrated the previous visualization Fraunhofer; TUM; TU Dortmund;
services with many partners in order to Universität Rostock (Germany); Everis;
compose various demonstration sce- UPV (Spain); KoçSistem; BOR; Defne;
narios on top of the Building as a Smartsoft (Turkey); X.com; Masaryk
Service platform, namely Smart University; and MDS (Czech Republic).
Booking Room, Maintenance applica-
tion, and Feedback application. Trials This research is also part of a horizontal
were carried out with end-users in a real task force with other ICT Future
environment, the Social Building of the Internet projects - such as FIWARE, in
Valencia City Council at Patraix. The particular FI-CONTENT 2 - that deals
facility manager, the room booking with building new innovative applica-
manager, teachers and citizens booking tions and services for every-day
and using the rooms for their own needs working and living environments, by
(foreign language classes, expositions, making ‘physical world information’
movies, presentations, yoga, etc.) were easily available for smart services. Our
given the chance to use the 3D web user work is partially supported by the
interface to visualize directly the con- Spanish MEC INNCORPORA-PTQ
figuration and sensing information of a 2011, MiTYC TSI-020400-2011-29,
room (smart booking room and mainte- and FEDER programs.
Advances in electronics and communication technologies are stimulating the growth of low cost connected
sensors and actuators. There are many potential application areas, for instance: home automation, security,
healthcare, smart grids, integrated transport systems, and next generation manufacturing. To date,
research in the area of the Internet of Things (IoT) has largely focused on the sensors, actuators and the
communication technologies needed for long battery life and efficient use of spectrum etc. rather than on
what’s needed to encourage applications and services. Current IoT products exist in isolated silos. To fully
realise the benefits, we will need scalable solutions for managing the vast of amounts of data, and open
standards that enable open markets of connected services.
The COMPOSE project is developing a obtained from the sensors, or passed to programming language and executed
open source cloud-based platform for the actuators associated with a given using the Chrome V8 engine.
IoT services. The run-time is highly scal- service object.
able and implemented on top of COMPOSE services are cloud-based.
CloudFoundry and Couchbase. This is Services can define composite streams Each service object is associated with a
complemented by the developer portal in terms of dynamic queries over other ‘web object’ at the network edge. The web
which supports discovery, registration, streams, for example, all temperature object provides an abstraction of the phys-
composition and deployment of services. sensors within 1.5 Km of a given loca- ical sensor or actuator, hiding the details
tion. Services can also define new of how these devices are connected. Web
COMPOSE models services in terms of streams as a transformation from objects communicate with the COM-
message streams, where each message another message stream, for example, to POSE platform using structured data rep-
corresponds to either a sensor reading or transform the physical units for the data resented with the JavaScript object nota-
an actuator update, and is structured as samples, or to smooth data samples. tion (JSON). There is a choice between
one or more data channels. Sensors and These mappings can be defined as using HTTP or WebSockets. The latter is
actuators are virtualised in the COM- simple expressions or more generally appropriate when the web object is behind
POSE platform as ‘service objects’. with scripts written in the JavaScript a firewall that prevents the COMPOSE
These define the message streams platform from opening HTTP connec-
Smart Environments, and in particular Smart Homes, have recently attracted the attention of many
researchers and industrial vendors. The proliferation of low-power sensing devices requires
integration gateways hiding the complexity of heterogeneous technologies. We propose a ZigBee
integration gateway to access and integrate low-power ZigBee devices.
Smart Environments, and in particular ZigBee offers a service-oriented Platform. ZB4O faces the problem of
Smart Homes, have recently gained the approach for low-power devices, with providing an easy access to low-power
attention of many researchers and hard- the unique feature of defining a variety sensing devices based on the ZigBee
ware vendors. An increasing number of of profiles that standardize the function- stack. It relies on the OSGi execution
sensing devices, whose price is rapidly alities of several classes of device (e.g., environment and it meets three basic
decreasing, are available on the market. home automation, health care, smart requirements: (i) it provides a rich and
Although such devices are becoming energy). OSGi offers a component- flexible gateway for the ZigBee net-
familiar in Smart Homes, user accept- based execution platform facilitating work, (ii) it extends the OSGi frame-
ance is limited by the fragmentation of the deployment and management of work with an open mechanism to inte-
the market. Heterogeneous technologies software units. grate ZigBee standard with a service-
do not integrate seamlessly into a Smart oriented approach and (iii) it defines an
Home; rather, each vendor offers its pri- The Wireless Network Laboratory at integration layer in order to access the
vate vertical solution. ZigBee and OSGi ISTI-CNR funds the ZB4O project [1], ZigBee network by using other tech-
play a predominant role in this scenario. ZigBee API for OSGi Service nologies.
Please contact:
Francesco Furfari and Michele
Girolami
ISTI-CNR, Italy
E-mail: {francesco.furfari,
michele.girolami}@isti.cnr.it
The Internet of Things (IoT) aims to interconnect large numbers of heterogeneous devices to provide
advanced applications that can improve our quality of life. The efficient interconnectivity of IoT
devices can be achieved with a hybrid Cloud Radio Access Network (Cloud-RAN) and Software
Defined Radio (SDR) framework that can overcome the heterogeneity of devices by seamlessly
adapting to their communication technology.
The Internet of Things (IoT) presents ronments where the ISM bands are ized server units that manage a pool of
itself as a promising set of technologies already overcrowded the interference base station resources, (ii) the distrib-
that enable the efficient and seamless from external sources (other wireless uted radio units (RUs) or SDR-based
interconnectivity of large numbers of networks) can severely impact network RUs (SRUs) that are located at remote
devices, allowing them to be discovered performance, since WSANs are very sites and create wireless cells providing
and accessed by the services in an susceptible to other wireless transmis- access to heterogeneous users and (iii) a
abstracted way. The IoT is considered to sions. But how can we allow the devices high bandwidth low-latency link
be one of the key technologies for 5G
communications and a basic technolog-
ical enabler towards the realization of
‘Smart Cities’. Smart Cities is a rela-
tively new term for a new generation of
cities that use new technology to drive
competitiveness, sustainability, eco-
nomic growth, energy efficiency and
improving Quality of Life (QoL).
relate with the users on each cell or cen- change frequencies when there is ture could greatly help mitigate the het-
trally by the centralized server units increased interference. The architecture erogeneity of the involved devices and
when interaction with the neighbour can also easily adapt to non-uniform their efficient coexistence. Since the
cells are required or for optimizing local traffic via the load balancing capability architecture is partly centralized, it can
decisions. An example of this architec- of the distributed pool of base stations. easily be integrated with existing IoT
ture is shown in Figure 1, where the dif- This pool can also share signalling, middleware platforms, playing the role
ferent colours of the SRU cells denote traffic and channel occupancy informa- of ‘communication manager’ - a func-
different communication technologies tion of the active users in order to opti- tional entity proposed in many IoT
and frequencies. mize the radio resource management architectures [3].
decisions. Spectral efficiency is
The software re-programmability of improved via the cognitive radio mech- This work has received funding from
SRUs enables any updates regarding the anisms for intelligent management of the European Union’s Seventh
communication protocols, installation spectrum resources that can be applied Framework Programme (FP7/2007-
of additional technologies, implementa- on the SRUs together with joint pro- 2013) under grant agreements no
tion of new networking standards, etc. cessing and scheduling. 609094 and 612361.
to be done easily, saving significant
costs associated with adding new hard- This architecture can have many advan- References:
ware at each remote site. The ability of tages for IoT, owing to its ability to opti- [1] “C-RAN The Road Towards Green
SDR to simultaneously handle different mize spectrum access for the multiple RAN”, China Mobile Research
communication technologies can facili- heterogeneous devices, prioritizing Institute White Paper Version 2.5, Oct
tate the virtualization of the remote access to network resources according 2011.
units in a way that each unit will be seen to the service request in both a central- [2] E.Z. Tragos, V. Angelakis:
as different Virtual Base Stations ized and a distributed manner. For “Cognitive radio inspired M2M
(VBS), each handling different commu- example, if two devices are in the same communications”, 16th International
nication technologies, e.g., VBS1 will area, even if they use different commu- Symposium on Wireless Personal
handle IEEE 802.11, VBS2 will handle nication technologies, they will Multimedia Communications
3G, while VBS3 will handle IEEE exchange their data directly through the (WPMC), pp.1,5, 2013.
802.15.4 – based IoT devices. local remote unit reducing unnecessary [3] A. Bassi, et. al: “Enabling Things
signalling. Another advantage is the to Talk”, Springer, ISBN: 978-3-642-
The centralized server unit is able to cost-efficient deployment of such an 40402-3 (Print) 978-3-642-40403-0
perform an optimized management of architecture, which only requires cen- (Online).
the available network resources since it tralized management and operation,
has a global view of the available while the installation of new remote Please contact:
resources at each SRU, and can recon- units can be done with a simple SDR Elias Z. Tragos
figure them easily at run-time to per- device and the required software. On FORTH-ICS, Greece
form, for instance, traffic offloading the road to 5G, where IoT is a basic E-mail: etragos@ics.forth.gr
when one unit is overloaded or to technological pillar, such an architec-
The world’s computing infrastructure is becoming increasingly differentiated into autonomous sub-
systems (e.g., IoT installations, clouds, VANETs), and these are often composed to generate value-added
functionality (systems of systems). But today, system composition is carried out in an ad-hoc, system-
specific, manner, with many associated disadvantages. We need a generalized system-of-systems-
oriented programming model that allows systems to be composed by application-domain ex-perts, not
just systems programmers. Furthermore, composition should occur in a principled way that generates
well-understood compositional semantics and behaviour.
Today’s distributed computing envi- tures such as smart cities and build- increasingly need to interact with and
ronment is becoming increasingly ings; environmental sensor and actu- respond to each other, typically in a
diverse and differentiated in nature, to ator networks using non-IP protocols; dynamic on-demand manner. For
the extent that the world’s computing cloud systems based on clusters; ad- example, WSNs need back-end clouds
infrastructure is now far removed from hoc networks such as MANETs and to process sense data, VANETs need to
the traditional picture of desktop PCs VANETs; and virtualized systems sup- interact with smart cities when upon
connected via fixed IP networks. The ported by network overlays. entry to the city, and overlay-based
picture today includes: IoT infrastruc- Furthermore, these various ‘systems’ systems need to maintain resilience
Recently proposed IoT platforms are typically based on centralized and cloud-centric
infrastructures. The BETaaS project aims at designing a platform for the execution of M2M
applications in a distributed runtime environment running close to the physical environment
where sensors and actuators are deployed. The main goal is to move the intelligence to the edge
in order to allow localized, content-centric, and timely processing of smart objects data. The
platform facilitates the integration of existing IoT systems, while providing software developers
with a high-level, content-centric abstraction to access smart objects’ resources.
Recent technology advances are cloud is generally inefficient in these use through the design and development of a
evolving the objects we use in our daily cases, and may be even unfeasible if the horizontal platform that leverages a dis-
lives into smart objects, i.e., regular case requires stringent latency guarantees tributed architecture to move the intelli-
objects empowered with communication to process and react to collected sensor gence to the edge, i.e., close to the phys-
and computational capabilities. Smart data [2]. ical environment where sensors and
home automation solutions (e.g., smart actuators reside, in order to ensure
thermostats, smart lights) and smart BETaaS, Building the Environment for timely application processing. The
health products (e.g., wearable sensors) the Things-as-a-Service, is a European BETaaS platform, is designed to run on
commercialized today are only a few project co-funded by the European heterogeneous devices, called gateways,
examples representing the dawn of the Commission under the 7th Framework such as home routers or set-top boxes.
forthcoming Internet of Things (IoT) Programme, which aims to overcome The execution of M2M applications is
revolution [1]. the limitations of cloud-centric IoT then supported through a tight interac-
architectures. This is accomplished tion among multiple gateways that form
Interoperability among heterogeneous
IoT systems is considered a key issue in
this scenario. Current IoT solutions are
typically vertical systems that are
designed to serve a specific application
acting in isolation with no or limited
cooperation. Instead, a horizontal
approach is needed to design future IoT
platforms in order to facilitate the cre-
ation of a converged infrastructure pro-
viding diverse applications a seamless
access to heterogeneous smart. This is
the clear trend also followed by ongoing
standardization activities related to IoT,
e.g. oneM2M.
Internet of things (IoT) and the related Industrial Internet are recognized as one of the most
significant technology-driven disruptions of the coming ten years. VTT Technical Research Centre
of Finland has chosen IoT as a strategic area for research, development and innovation projects.
The Pro-IoT spearhead programme research centre. Each area is delineated to point out the relevance of the pro-
focuses on three technology areas, in three domains that represent chal- posed solutions for companies. The
namely sensing technology, connec- lenges for our societies: industrial asset complete structure of the programme is
tivity, and data-analysis with data secu- management, connected health and dig- presented Figure 1.
rity. These areas emanate from the needs ital society. In addition to the technolog-
expressed by European and especially ical perspectives, the business aspects In order to sustain living standards in
Finnish companies, as well as on the are also investigated, through a dedi- Europe, an increase in the productivity
strong competencies found in our cated module, hence making it possible of companies appears to be vital. As
Research Highlights
The development of the Pro-IoT pro-
gramme has resulted in many research
projects. These projects have yielded
significant achievements in various
domains. A notable example, in the
field of sensing, concerns wireless and
battery-less devices, which represent a
kind of Holy Grail. Such a paradigm Figure 1: VTT productivity with IOT program structure.
would enable measurements in
demanding conditions with low main-
tenance cost and without hindering pro- manage the multi-technology context. intensified research effort involving the
duction. VTT has developed a solution Therefore, a platform has been devel- key stakeholders. We need to further
for measuring temperature, inclination, oped for energy-efficient communica- address technological questions such as
humidity, strain and other characteris- tions between mobile devices that the need of more energy-efficient com-
tics from machines and inside struc- identifies the most appropriate combi- munication, the congestion of the
tures (e.g. walls) at distances up to 10 nation of technologies for different Internet with the deployment of tens of
m, without batteries in the sensor. The exchanges [1]. billions of new nodes (sensors and
principle is based on powering the smart objects). In addition, we have to
sensor with radio frequency energy Impact on Industry and Society find sustainable answers to the recurrent
waves and using the return transmis- The disruption caused by Industrial questions of information security and
sion to read out the resulting data (see Internet is comparable to earlier indus- privacy in this new context. The safety
Link ‘Zero Power Sensor’). The IPR trial revolutions initiated by steam, aspect of the operations regarding
protected technology allows for several electricity and computers. Indeed, autonomous machines, the issues of
individually identifiable sensors to be General Electric predicts extra growth data ownership, as well as plausible and
operated in the same space. globally worth 10 to 15 trillion US dol- fair business models also represent
lars during the next 15 years. This will major concerns.
Another relevant example concerns depend largely on being able to
autonomous vehicles and machines leverage industrial internet technology Links:
with communication needs. A solution to the full [2]. As such, the Pro-IoT pro- The Pro-IoT spearhead project:
for communication in demanding con- gramme contributes to this evolution http://www.vttresearch.com/impact/new
ditions, namely a cluttered harbour by providing companies with appro- -innovations-are-here/innovation-
environment, has been developed for priate tools to improve the performance programmes/productivity-with-
and tested in Port of Singapore. The key of their industrial processes. As shown internet-of-things
issue was how to ensure reliable com- by the examples in the previous sec- Zero Power Sensor:
munication with predictable maximum tion, the emphasis is on realistic solu- http://knowledge.vtt.fi/ZeroPowerSensor
delay between moving machines for tions adapted to a challenging environ-
their safe operation. The solution ment, safe autonomous systems while References:
involved using two or more communi- targeting a reduction in the operating [1] S. Chaumette and J. Ouoba: “A
cation networks seamlessly to secure costs. Multilevel Platform for Secure
communication between piles of con- Communications in a Fleet of Mobile
tainers and other ‘non-radio-friendly’ The Industrial Internet will not only Phones”, in proc. of MobiCASE 2014
objects. increase productivity in industry and [2] P. Evans and M. Annunziata:
firms. By adapting its models to urban “Industrial Internet: Pushing the
At a more conceptual level, other areas, it will also affect public services Boundaries of Minds and Machines”,
research activities were intended to with the emergence and the consolida- 2012, http://www.ge.com/docs/
address the issues related to the multi- tion of smart cities. The everyday lives chapters/Industrial_Internet.pdf
technology environment of mobile of all of us as employees, consumers
devices. Indeed, many of these devices and citizens will benefit from it with Please contact:
(mobile phones, sensors, wearable more personalized and efficient end- Ailisto Heikki
objects) are endowed with multiple user services. VTT Technical Research Centre of
wireless technologies. This offers Finland
opportunities in terms of services and Challenges Ahead E-mail heikki.ailisto@vtt.fi
applications to deploy, provided that The Industrial Internet is not happening
solutions are designed to effectively by itself; it requires coordinated and
A new Horizon 2020 Innovation Action project fosters the uptake of music technology research and
acts as an accelerator for bringing creative music tech ideas to market, while providing a mesh of
technological ‘bricks’ to create an ‘Internet of Music Things’.
Music is arguably the most pervasive of and little connectivity hinder the cre- bass transcription from Fraunhofer,
the performing arts. It has the power to ation of larger systems that operate as a melody curve extraction and API access
enlighten us, cheer us up, bring us down mesh of technological bricks con- to the freesound.org sample collection
and most importantly - bring us together. necting physical and tangible hardware by Universitat Pompeu Fabra. On top of
Substantial resources and personal and available music technology soft- these APIs, graphical interfaces (GUIs)
energy have been invested into music ware. Yet this interoperability between such as PlaySOM and Sonarflow offer
technology research in recent decades. tangible and wearable devices and the semi-automated grouping and clus-
Despite the global success of European well-researched software components tering of music while other visual fron-
music tech companies such as Spotify, for music analysis and processing is tends shall be easily used as well in con-
Deezer, Soundcloud and last.fm, the heavily needed to make the ‘Internet of junction with the tools and APIs. The
value chain from academic research to Music Things’ a reality [1]. third category of bricks is a new genera-
SMEs and large music industry corpora- tion of Tangible User Interface (TUI)
tions remains fragmented, preventing The MusicBricks Project such as a wireless micro platform for
successful application of the results of MusicBricks responds to these prob- motion sensing provided by IRCAM [2]
innovative research. Yet, creative SMEs lems, acting as a connector between This is complemented by portable com-
have been identified as the primary cata- research and industry, by identifying, puting platforms such as the Arduino,
lysts for innovation, and they can benefit wrapping and making existing tools the Raspberry Pi and the Axoloti, a
vastly from innovative music tech- available in easily usable formats. microcontroller specifically designed to
nology research. These tools, or ‘bricks’, shall require create free-form digital audio instru-
little learning, offer themselves for ments of a new kind.
The Horizon 2020 Innovation Action mash-ups, and are easily adopted and
‘MusicBricks’ has been initiated with deployable by the creative SME digital The Internet of Music Things
the aim of capitalizing on the current makers and content creators. The All these technological ‘bricks’ form the
European leadership in successful music project will foster interoperability by foundation for a co-creational space
technology companies and secure a providing state-of-the art connectors where people - musicians, hackers, dig-
direct route from ICT academic research and interfaces to these technological ital makers, creative SMEs - build
to the application and exploitation by building blocks. The resulting plethora entirely new applications. By intercon-
SMEs and major industry players world- of components - software endpoints, necting individual nodes - the provided
wide. The mission of the project is to Web APIs, physical, tangible and wear- API endpoints, network-enabled
foster an exchange between the music able devices - are connected through devices, sensors, microcomputers etc. -
ICT experts from academic and national and live on the Internet, realizing an these bricks communicate with each
research organizations, and digital ‘Internet of Music Things’, where many other and start creating something
makers and content creators from cre- different kinds of analysis, processing, bigger: a mesh of connected ‘musical
ative industries SMEs. sensing, actuating or synthesis are components’ which eventually manifests
taking place on different machines. as an ‘Internet of Music Things’.
Despite the remarkable progress in
music information research, technolo- The consortium consists of the design Events such as Hack days and Music
gies provided by various projects are research and innovation lab Tech Fests fuel the creation of new
often difficult to apply owing to poor Stromatolite, the Institut de Recherche mash-ups, hacks, audio tools, music
communication. Research results are et Coordination Acoustique/Musique instruments combined with unthought-
typically scattered among institution (IRCAM), the Music Technology of forms of application which has been
Web pages and documentation is often Group of the Universitat Pompeu Fabra, demonstrated impressively in the past
neglected. Less user-friendly interfaces the Fraunhofer IDMT Institute, the by innovations such as Siftables [3],
and steep learning curves hinder uptake Vienna University of Technology, and is compact devices with sensing, graph-
by industry (even more so for SMEs led by Sigma Orionis. ical display, and wireless communica-
and, in particular, one-person enterprises tion that can be physically manipulated,
such as the typical music maker or app In the first step the project will provide recognize gestures and sense other
developer). In addition, corresponding the ‘bricks’ for creative makers: soft- nearby devices while able to interac-
technologies are frequently lacking ware tools and APIs, such as rhythm tively create audio; or performances
proper engineering to facilitate interop- feature analysis and music similarity by such as the Brainwave Quartet in which
erability with other technological com- Vienna University of Technology, key music is performed live through the col-
ponents. Thus, complicated workflows and tempo detection and melody and lective brainwaves of an ensemble
Links:
Music Bricks Website:
http://musictechfest.org/MusicBricks
Music Tech Fest:
http://musictechfest.org
Results of previous Music Tech Fests:
http://musictechfest.tumblr.com
Videos from Music Tech Fests:
https://www.youtube.com/user/Music
TechFest
Reference:
[1] Music’s Internet of Things:
wearing brain caps. Another great Through this interdisciplinarity it Heartbeats, Accelerometers...
example is the “Wearable Axoloti enables a new dimension of creativity, Brainwaves?
Music Human Synthesizer”, a multi- leading to entirely novel ideas and http://evolver.fm/2013/03/15/musics-
user synthesizer where one user modu- applications. In past events creative internet-of-things-heartbeats-
lates the sound of another through hacker ideas have led to physical prod- accelerometers-brainwaves/
touching, created with wearable elec- ucts and performances featuring new [2] Jules Françoise, Norbert Schnell,
tronics, conductive surfaces on T-shirts musical innovations such as a ‘Music Riccardo Borghesi, and Frédéric
and Axoloti boards. It was awarded the Hat’ which converts brainwaves and Bevilacqua.Probabilistic Models for
‘Internet of Music Things Hack Award’ head movements to music and various Designing Motion and Sound
at the Paris Music Tech Fest. effects. Relationships. In Proceedings of the
2014 International Conference on New
Events The most promising ideas and demon- Interfaces for Musical Expression,
MusicBricks both hosts and cooperates strators generated at these events - NIME’14, London, UK, 2014.
with a series of events, specifically the whether a piece of software, a device, a [3] D. Merrill, J. Kalanithi and P.
Music Tech Fest geared towards the cre- new musical instrument, a performance Maes. Siftables: Towards Sensor
ative developer community to actively or installation - will be supported by the Network User Interfaces. In the
test and utilize these tools, alongside MusicBricks Incubation Programme, Proceedings of the First International
others, to generate new ideas and create which provides funding for residencies Conference on Tangible and Embedded
novel applications. The Music Tech that enable creative makers to further Interaction (TEI'07). February 15-17 in
Fest (Figure 1) has seen successful past develop their demonstrators to robust Baton Rouge, Louisiana, USA.
events around the world and is a free, and market-ready prototypes and assists
weekend-long event that provides an them through virtual and face-to-face Please contact:
experimental and improvisational space collaboration. Feedback gathered Thomas Lidy, Alexander Schindler
where ideas are showcased, performed, during idea generation and incubation Vienna University of Technology
demonstrated and discussed, while will assist in the consolidation and lidy@ifs.tuwien.ac.at,
offering immediate assistance with the refinement of the available tools and schindler@ifs.tuwien.ac.at
tools provided. It has a particular focus interfaces. By exposing the resulting
on interdisciplinarity: musicians meet prototypes directly to big industry Michela Magas
hackers, researchers meet industry, players and investors the project aims to Stromatolite
artists and technologists come together. reach the global market and gather michela@stromatolite.com
NFC tags and transceivers are ubiquitous and well supported. Like many academic research groups,
the Communication Systems Group CSG of the University of Zürich owns many physical devices,
which are required for research and teaching. Traditionally, a printed, human-readable inventory and
attached labels have been used to keep track of these devices. A new inventory approach was
developed with the aim of simplifying the data acquisition using an NFC tag-based system supported
by an Android application. This approach was implemented, tested, and used productively. However,
due to technical difficulties - namely, a poor response of NFC tags on metallic material - the NFC-
based inventory system did not simplify data acquisition and consequently, the NFC-based inventory
system was changed back to the label-based system.
Near Field Communication (NFC) To avoid not only the need for a special- Here, NFC tags qualified as the only
defines a communication channel for ized printer but to also allow every option, as other RFID tags would
devices in close proximity (up to 10 cm) group member to conveniently retrieve require extra hardware for reading and
[1]. NFC has a wide range of applica- and change inventory information, a writing tags. This would not only
tions as it can be used to transmit data new tag-based inventory system was increase costs of the new approach but
between devices or between a device and envisioned. Barcodes, QR-codes, and also reduce user convenience. Thus,
an NFC tag. These tags are standardized RFID tags were evaluated to label items. since NFC is already supported by many
– though some differences exist – and the Bar- and QR-codes were not followed smartphones and tablets, the NFC tags
data is stored in the standardized format up for two reasons: first, creating determined a valuable path.
NDEF (NFC Data Exchange Format) human-readable tags would again imply
[2]. Depending on the type of tag, data the need for a specialized label printer, Thus, the human readable inventory
from 48 Bytes (Ultralight) to 888 Bytes which would prohibit “decentralized” labels were replaced with NFC tags, to
(NTAG216) can be stored. The very pop- allow receipt and updating of inventory
ular tag NTAG203 and its successor information upon scanning the tag with
NTAG213 can store 144 Bytes of data. an NFC-enabled smartphone. To add a
Most Android-based smartphones are new device a database entry is created
equipped with an NFC transceiver, and via the newly developed inventory
the software support for NFC has been smartphone app or via the inventory
present since the Android version 2.3, Web site. An NFC NTAG203 tag is
which was released in December 2010. placed on the device and an NFC-
Apple also integrated the NFC tech- capable device used to store this entry’s
nology into iPhone 6 and bases its identifier and other essential informa-
mobile payment system on this tech- tion in those tags 144 Byte storage.
nology, although not providing an open Updating entries is simple: an NFC tag
API. Thus, NFC reading and writing is scanned with the inventory app and all
capable devices are widespread and available information about this item is
ubiquitous. displayed immediately. This informa-
tion (including location, room, and
Inventory System holder) can be updated within the app,
The label-based Communication and written to the tag and the database.
Systems Group CSG [3] inventory
system for hardware acquisitions above a Figure 1: App-based inventory system. The smartphone inventory app was
certain commercial value used a human implemented in Java, while the backend
readable, four character and four digit was implemented in PHP communi-
label. The label had to be printed with a tagging of devices. Second, once these cating via HTTPS requests to the smart-
specialized label printer and was tags are created, information they store phone app. Authentication is done with
attached to the device. The four digit is static. This was a major drawback, LDAP, while all other data is stored in a
number was used to identify a database because the new inventory system has to MySQL database. Updating of device
entry, which contained information, such be able to operate offline, i.e., even holders is the most frequently performed
as device name, vendor, purchase date, without a database connection retrieving action, and the app’s starting screen (cf.
and current holder. Retrieving inventory or updating inventory information of a Figure 1) offers two buttons for this pur-
information about a device involved device in physical reach had to be pos- pose. Upon selecting ‘Lend item’, the
manually entering its number into a Web sible. Therefore, change of information name of the new item holder is
browser. on a tag and its synchronization later requested. Once entered, an NFC chip
with the database was essential. can be scanned, which, like the
Consequently, RFID tags had to be used. according data base entry, is updated
accordingly. The ‘Return’ button deletes shielded tags ranging in price from a Once metal-cased devices work well
the ‘current holder’ field of the subse- few cents to a few dollars. The cheap with a new type of metal-shielded tags,
quently scanned item (in the data base metal shielded tags performed as poorly the inventory system will be switched
and on the NFC chip). ‘Manipulate as non-shielded tags, whilst the more back to an NFC-based inventory
item’ allows a database entry to be dis- expensive ones generally performed system. In conclusion, when planning
played and modified when the corre- better than the low-cost tags. However, an introduction of an NFC-based
sponding item is scanned. Thus, items the evaluations have shown that even system, it is important to ensure that
can also be added or deleted here. With a the most expensive and metal-shielded NFC tags on metallic items can be used
‘More’ click, additional information and tags failed to work on a large set of by verifying NFC tags operations on
a search function are displayed. metallic items. metallic items.
The TrakChain assessment tools take a description of a physical supply chain – relevant locations, how
many goods are received, how often, etc. – and estimate the performance of track and trace queries in
a modelled traceability system, providing predictions of how much processing and storage will be
required for the working system. The tools were developed at Instituto Superior Técnico, Universidade de
Lisboa, Portugal and were evaluated using a Pharmaceuticals supply chain case study.
The Internet of Things (IoT) promises However, to achieve further improve- locations generate event data, as illus-
benefits from a deeper connection ments, more up-to-date and precise trated in Figure 1.
between the virtual and physical worlds. information about the supply chain is
One specific application area is logis- required. A practical RFID traceability system
tics. The global economy depends on a should perform adequately for the large
wide range of supply chains that transfer RFID is an IoT technology that allows number of physical objects flowing in
goods from producers to consumers. detailed and automated data capture in the supply chain [1]; and it should pro-
The combined use of Enterprise the supply chain, as specified by the tect the sensitive business data from
Resources Planning (ERP) and Supply EPCglobal standards [2]. Tags are unauthorized access providing the
Chain Management (SCM) information attached to the objects of interest and desired data visibility [2]. The
systems has greatly improved the overall readers placed along the supply chain TrakChain project was proposed to
operational efficiency of supply chains. evaluate both these aspects. It provides
parameterise
Please contact:
Supply chain compute
Miguel L. Pardal
schema Cost model Estimates Instituto Superior Técnico, Portugal
Email:
Figure 2: Data flow of the TrakChain cost assessment tool. Miguel.Pardal@tecnico.ulisboa.pt
Mesh Joinery:
A Method for Building
Fabricable Structures
by Paolo Cignoni, Nico Pietroni, Luigi Malomo
and Roberto Scopigno
Link:
http://vcg.isti.cnr.it/Publications/2014/CPMS14/
References:
[1] P. Cignoni, N. Pietroni, L. Malomo, R. Scopigno:
“Field-aligned mesh joinery”,. ACM Trans. Graph, 33, 1,
(February 2014).
[2] Y. Schwartzburg , M. Pauly: “Design and optimization
of orthogonally intersecting planar surfaces”, Computation-
al Design Modelling, 2012.
[3] D. Panozzo, Y. Lipman, E. Puppo, D. Zorin: “Fields on
symmetric surfaces”, ACM Trans. Graph. 31, 4 (July), 2012.
Please contact:
Paolo Cignoni
ISTI-CNR, Italy
E-mail: paolo.cignoni@isti.cnr.it
Icing detection and mechanism. Very promising results come from carbon nan-
otechnology that can be used to make thin films of coating
Protection for Small material to be painted on aircraft wings. Using the electricity
onboard, the film can be heated to melt the ice. Such devices,
Unmanned Aircrafts however, are very energy demanding, thus it is vital to have
an optimal scheme for the activation of the de-icing system
by Andrea Cristofaro and Tor Arne Johansen to maintain aircraft stability. The combination of the aircraft
dynamical model and the model describing the accretion of
In unmanned aerial vehicles, it is vital to be able to icing is the ‘Stefan problem’, which consists of a set of four
detect and accommodate ice adhesion on wings, partial differential equations: two heat equations, one mass
control surfaces and airspeed sensors, since ice balance equation and one energy balance equation.
accretion modifies the shape of the aircraft and alters
its measurements, thus changing the aerodynamic Finally, in order to test and validate the efficiency and robust-
forces and reducing manoeuvring capability. ness of the icing detection algorithms as well as the icing
protection module, field experiments in Svalbard are planned
The use of unmanned aerial vehicles (UAVs) for civil appli- for the near future using the X8 UAV (Skywalker
cations – largely surveillance and monitoring - has increased Technology Co, Ltd.).
markedly in recent years. UAVs are very often suited for
harsh conditions that are unsafe for humans, such as those The research at NTNU is a fruitful collaboration with other
typically encountered in Arctic operations. Reliable, appro- two ERCIM partner institutes, namely University of Porto
priate, and efficient UAV operations are needed in such con- (UPT) and University of Cyprus (UCY). In particular, we are
ditions. currently working jointly with Prof. Pedro Aguiar (UPT) and
Prof. Marios Polycarpou (UCY) who have made outstanding
The phenomenon of ice accretion on aircraft wings and con- contributions to the project with their expertise in multiple
trol surfaces is a well recognized problem in aerospace engi- model adaptive estimation and nonlinear fault detection,
neering: the modified shape of the leading edge due to ice, respectively.
changes the lift, drag and pitch moment characteristics of the
wing. For instance, an ice-covered airfoil may experience a
40% reduction in lift coefficient, while the drag may be
increased by as much as 200%. A decrease in lift requires
more engine power and implies a premature airfoil stall
angle.
A second step towards achieving icing protection for UAVs, Tor Arne Johansen, NTNU, Norway
is the implementation of an efficient anti-icing and de-icing E-mail: tor.arne.johansen@itk.ntnu.no
When displaying the knowledge graph, users have different Please contact:
options. In the basic view, documents are structured Ulrich Nütten, Fraunhofer IAIS, Germany
according semantic similarities and shown as interlinked Tel: +49-2241-142915
nodes. Node colours reflect subject areas and the most general E-mail: ulrich.nuetten@iais.fraunhofer.de
Online Semantic Analysis Finally, the third goal of our approach is that the method
should operate at (close to) wire speed. Even when real-time
over Big Network data analysis is not strictly needed, the off-line method is limited
by storage because the analysis capacity and capabilities
by Chengchen Hu and Yuming Jiang cannot keep up with the rate at which data is produced.
Extracting and storing only useful information is a viable
Mobile applications, web services and social media approach that needs to be explored further.
generate huge data traffic from many sources. This is
described as ‘big network data’ with the ‘4V- In order to achieve the above goals, we apply Deep Semantic
characteristics’ - that is, variety, velocity, volume and Inspection (DSI), which contains a standard description to
veracity. Analysis on such big network data can discover unify the various formats of different applications and finally
problems, reveal opportunities and provide advice for obtain user semantics. Our basic idea is to extract a minimum
the service and network providers for fine-grained but complete semantic for each user behaviour at wire speed,
traffic engineering with close to real-time adjustment or and then apply data analysis and data mining on the small
application acceleration [1]. sized semantic data instead of the raw traffic data. This
process purifies the raw traffic and reduces the data volume
Different applications tend to integrate several functionali- by several orders of magnitude. Our preliminary experiment
ties with various data formats. For example, the Twitter shows that the compression ratio between the raw traffic
application produces network traffic such as tweeting, volume and our approach is in three orders of magnitude. As
posting pictures, embedding video. Classical methods are a result, the data volume for further user-defined high-level
limited to protocol or application identification [2]. We need analyses can be significantly reduced to handle big, and
to go beyond the packet and application analysis, when increasing, network data.
semantic information is the target [3]. Therefore, the first
goal is for our method to exhibit fine-grained awareness, We have designed and implemented a cross-platform system,
which analyzes user behaviour instead of traffic only related named Semantic On-Line Intent Detection (SOLID) to
to a certain application. This implies that we need to use a realize our DSI approach. As shown in Figure 1, SOLID
general grammar to associate the unstructured data with user builds a processing pipeline in its kernel space, where a
behaviour. semantic parser translates the segments into the application
semantic. A matching engine compares the application
The second goal is to develop a flexible and uniform specifi- semantic with predefined specifications to output the user
cation of the user semantic from network traffic. In previous sketches. The kernel design allows simultaneous processing
work, heterogeneous and unstructured big network data in of multiple application specifications and multiple PDUs.
different formats are studied separately. It is a challenge to Our implementation has demonstrated that on a real traffic
normalize the independent data structures and describe user trace, the SOLID-system achieves 17.6Gbps throughput with
behaviour in a unified framework to conduct a comprehen- the cost of 709MB memory consummations in x86 platform.
sive analysis in a fine-grained semantic manner. To make the SOLID-system flexible so that it supports var-
Please contact:
Bruno Michel
IBM Zurich Research Lab, Switzerland
E-mail: bmi@zurich.ibm.com
VAVID allows technology firms to get a better grip on the The data being studied by the project are numerical simula-
massive amount of data they need to handle. The partners in tion results data from the automotive and wind industries as
this project are developing methods to tackle the enormous well as measurement data taken from wind turbine moni-
volumes of data that accumulate at engineering departments. toring systems. By performing joint and comparative
Examples of such data include simulation results and the analysis of data from different industries, the partners are in
sensor data received from machines and installations. VAVID the first project phase developing a methodology for efficient
works by using comparative analysis and data compression to data analysis. These methods and techniques are going into
reduce data to its relevant core. This saves on the costs of data the creation of a high-performance data management system
storage and creates the transparency needed by engineers to that will allow centralized data storage as well as efficient
optimize both production and products. VAVID is being coor- data access and retrieval.
dinated by the Fraunhofer Institute for Algorithms and
Scientific Computing (SCAI) and is receiving 2.2 million € in The second phase will focus on the end user by developing
support from the German Ministry of Education and Research innovative, universally applicable software components that
(BMBF) under its Big Data program. can later be deployed in other industries. Examples of note
here include the fields of aviation and mechanical and plant
engineering (industrial plant).
More information:
http://ronchi.isti.cnr.it/AITA2015/
More information:
http://icec2015.idi.ntnu.no/
FWO F.R.S.-FNRS
Egmontstraat 5 rue d’Egmont 5 Magyar Tudományos Akadémia
B-1000 Brussels, Belgium B-1000 Brussels, Belgium Számítástechnikai és Automatizálási Kutató Intézet
http://www.fwo.be/ http://www.fnrs.be/ P.O. Box 63, H-1518 Budapest, Hungary
http://www.sztaki.hu/
University of Cyprus
Foundation for Research and Technology - Hellas P.O. Box 20537
Institute of Computer Science 1678 Nicosia, Cyprus
P.O. Box 1385, GR-71110 Heraklion, Crete, Greece http://www.cs.ucy.ac.cy/
FORTH http://www.ics.forth.gr/
University of Geneva
Centre Universitaire d’Informatique
Battelle Bat. A, 7 rte de Drize, CH-1227 Carouge
Fraunhofer ICT Group
http://cui.unige.ch
Anna-Louisa-Karsch-Str. 2
10178 Berlin, Germany
http://www.iuk.fraunhofer.de/
University of Southampton
University Road
Southampton SO17 1BJ, United Kingdom
http://www.southampton.ac.uk/
INESC
c/o INESC Porto, Campus da FEUP,
Rua Dr. Roberto Frias, nº 378,
4200-465 Porto, Portugal Universty of Warsaw
Faculty of Mathematics, Informatics and Mechanics
Banacha 2, 02-097 Warsaw, Poland
http://www.mimuw.edu.pl/