=Paper=
{{Paper
|id=Vol-1641/paper5
|storemode=property
|title=A Conceptual Framework For End-User Development in Ambient Intelligence
|pdfUrl=https://ceur-ws.org/Vol-1641/paper5.pdf
|volume=Vol-1641
|authors=Federico Cabitza,Daniela Fogli,Rosa Lanzilotti,Antonio Piccinno
|dblpUrl=https://dblp.org/rec/conf/iseud/CabitzaFLP15
}}
==A Conceptual Framework For End-User Development in Ambient Intelligence==
https://ceur-ws.org/Vol-1641/paper5.pdf
A Conceptual Framework For
End-User Development in Ambient Intelligence
Federico Cabitza1, Daniela Fogli2, Rosa Lanzilotti3, Antonio Piccinno3
1Dipartimento di Informatica, Università degli Studi di Milano-Bicocca, Milano, Italy
cabitza@disco.unimib.it
2Dipartimento di Ingegneria dell’Informazione, Università degli Studi di Brescia, Brescia, Italy
daniela.fogli@unibs.it
3Dipartimento di Informatica, Università degli Studi di Bari Aldo Moro, Bari, Italy
{rosa.lanzilotti, antonio.piccinno}@uniba.it
Abstract. This paper proposes a conceptual framework for the design and con-
tinuous evolution of ambient intelligence environments. The framework is based
on the metaphor of interconnection between humans and devices belonging to
the intelligent environment. All actors are called on to perceive and act on the
environment itself, giving rise to novel and unanticipated behaviors. To this aim,
the proposed framework is structured along three layers - physical, inference and
user - sharing an information space of events, conditions and actions. In particu-
lar, a EUD approach is advocated for the user layer to support ambient configu-
ration and management and foster a smooth and progressive co-evolution of users
and technologies.
Keywords: end-user development, interconnection, inference rules, socio-tech-
nical system
1 Introduction
The Ambient Intelligence (AmI) field is aimed at realizing scenarios where people will
move and live in “intelligent” electronic environments, that is, in active places that have
claims of being sensitive and responsive to their needs [1]. The choice of the adjective
“ambient” is not fortuitous: it refers clearly to the fact that these electronic environ-
ments are not just “out there”, as simply given to the human experience. Rather, they
are something that surrounds and encircles users (i.e., inhabitants) and is formed both
by their perceptions (in that the system modulates and filters away information accord-
ing to the context in order to meet users’ needs) and by their activities (in that users’
actions are an important part of the context the intelligent environment is sensitive to).
Therefore, an AmI includes and cannot exist without the intelligence of the users, i.e.
without the “human” sensors and actuators making the system “sentient” [2].
This view brings into conceiving an AmI environment as a socio-technical system,
which encompasses not only a variety of hardware and software components, such as
sensors, robots, actuators, apps and Web applications, but also people that are bound
together by social ties and personal relations of acquaintance and that are linked each
27
Proc. of Third International Workshop on Cultures of Participation in the Digital Age - CoPDA 2015
Madrid (Spain), May 26th, 2015 (published at http://ceur-ws.org).
Copyright © 2014 for the individual papers by the papers' authors. Copying permitted for private and academic purposes.
This volume is published and copyrighted by its editors.
other through their personal devices and the hardware/software components distributed
in the environment. This results in a socio-technical network of humans and non-hu-
mans, where both kinds of actors are able to perceive and act in the environment. An
important feature of such a network is the ability to cope with the continuous changes
in the environment and the network itself, by reconfiguring human-human, machine-
machine and human-machine relationships [2]. The ways in which each user configures
the ambient surrounding her/him have deep consequences every time this environment
is shared by multiple users, as also their preferences, needs and expectations come to
be shared and get intertwined in the same environment, possibly leading to conflicts or
incoherent behaviors. Consequently, we advocate the adoption of an End-User Devel-
opment (EUD) approach [3] in the design and continuous evolution of this class of
systems. EUD techniques would support end users in shaping the environment where
they live by means of tools and interaction metaphors that are suitable to people without
specific competencies in software technologies and programming languages.
EUD in ambient intelligence has not received, till now, the necessary attention. The
survey reported in [4] underlines that many studies carried out in AmI focus on one
hand on hardware devices (sensors and actuators) and, on the other hand, on artificial
intelligence techniques that may help to add sophisticated capabilities to the processing
of information provided by hardware devices. Among end-user programming tools,
there are some interesting proposals that are mainly focused on single-user interaction:
one single user is called on to define the behavior of his/her smart things [5]. Interac-
tions among things and between humans and things within and intelligent environment
are not usually considered, thus remaining far from satisfying the challenges and design
principles identified in [6], [7]. Literature reveals that the Event-Condition-Action
(ECA) rule-based paradigm is the most promising approach for the development of the
behavior of an intelligent environment [8], [9]. The intuitiveness of the ECA paradigm
has been demonstrated both in the ubiquitous computing field [10] and in other appli-
cation domains (e.g. [8]). Recent web services, which allow users creating their own
“recipes” for controlling their environment through their personal devices, are based on
the same paradigm: see for example IFTTT1 and Atooma2. Furthermore, the filtered
lists metaphor adopted in these systems seems to be a good compromise between ex-
pressiveness and simplicity, thus favoring end-user engagement and participation.
In this paper we delineate a conceptual framework for the development, maintenance
and evolution of a diffusely intelligent and cooperative socio-technical environment.
The idea underlying this framework is that of interconnection [11], so as to go beyond
the concept of “single device programming” in favor of the collective and participatory
development of an intelligent system for the sake of a community of users who could
directly or indirectly intervene in such a development. In particular, we focus on the
EUD features of the framework. Open issues and possible research directions are also
finally discussed.
1 www.ifttt.com
2 www.atooma.com
28
Proc. of Third International Workshop on Cultures of Participation in the Digital Age - CoPDA 2015
Madrid (Spain), May 26th, 2015 (published at http://ceur-ws.org).
Copyright © 2014 for the individual papers by the papers' authors. Copying permitted for private and academic purposes.
This volume is published and copyrighted by its editors.
2 The Conceptual Framework
We propose a framework based on the central concept of interconnection among all
actors involved in the intelligent environment, i.e., devices and humans, and of fluid
exchange of states, conditions and behaviors. This should be guaranteed so that the
intelligent environment could really align dynamically and adaptively to the user needs
of information and support for action. The main point of this metaphor is that an AmI
environment is not simply an aggregation of computational devices endowed with ei-
ther perceptual or effectual capabilities, or both of these, bound together by functional
relations; but rather what emerges as something new from this aggregation. In other
words, it is an environment where some states and conditions are visible and available
(e.g., the temperature of the room, what time it is, if the door is open, etc.) and some
actions are feasible (e.g., raise the room temperature, mute the mobile, call the security,
etc.). And, last but not the least, where “intelligent inference” is kept in the loop be-
tween conditions and actions, so that some value for the ambient inhabitants is created
through a collective intelligence [12]. The point is that value is created by interconnec-
tion, that is by having any cognitive or computational actor getting access to the whole
information space made available (and also created on purpose) so that proper action
can be performed both by the system supporting the intelligent environment and the
inhabitants at due time. Like in a solution where chemical reactions happen according
to the actual elements poured in, and not necessarily according to a master plan that
precedes the event of mixture, so should an AmI environment not be designed a priori,
but rather be able to adapt over time. To some extent we advocate that an approach
similar to that conceived for collaborative applications with the phrase “design for un-
anticipated use” [13] is adopted also for AmI systems.
The above considerations bring to conceive an AmI environment as emerging from
the good fit of three layers into a single coherent conceptual architecture (see Figure 1):
1) the physical layer, where devices operate, with their perceptual and effectual capa-
bilities, and inter-operate, through a shared information space that collects their states
and acts as both a blackboard and a communication medium; 2) the inference layer,
where logical reasoning is automatically performed to link perceptions to actions; 3)
the user layer, where people are called on to express their preferences through actions,
also in response to state changes and/or ambient conditions, as well as to modify and
manipulate the “logic” (i.e., rules, policies, complex behaviors) their ambient should
follow to either assist or satisfy them (EUD activities).
The last 10 years or so of AmI research has brought to devices that are sufficiently
reliable and capable to interoperate with other devices and the overall hosting infra-
structure [4]. Also the inference layer has reached a maturity level where consolidated
inference engines combine efficiency with flexibility (e.g., [14]). Therefore, time has
also come to invest more efforts in the user-related layer. This means to address new
challenges like the perception of relative advantage and long-term sustainability. The
former regards a known principle from the diffusion of innovation model theory [15]:
it is important that users perceive the new thing as giving them a clear advantage with
respect to the traditional counterparts (whatever these are) in order to have an impact
on their daily life. For example, in the AmI context, this means that setting up a home
29
Proc. of Third International Workshop on Cultures of Participation in the Digital Age - CoPDA 2015
Madrid (Spain), May 26th, 2015 (published at http://ceur-ws.org).
Copyright © 2014 for the individual papers by the papers' authors. Copying permitted for private and academic purposes.
This volume is published and copyrighted by its editors.
lightening policy should be somehow as easy as switching a light on manually (or
easier!). In order words, end users should consider participation in rule creation as a
meaningful activity and not as an additional duty that they must satisfy. On the other
side, the sustainability challenge is closely related and regards the fact that users
should find convenient to interact with their intelligent “ambients” even after that the
thrill for the technology novelty and curiosity have faded out and/or after that some
expert has gone away and left the lay user alone with her/his own computational sup-
port. This is an aspect that is seldom considered in the AmI literature and it means to
carefully design the user layer and its EUD features, so that an AmI environment can
smoothly co-evolve with the users’ needs and technologies available (see Figure 1),
without forcing users to perform difficult and undesired tasks.
Fig. 1. The conceptual framework.
3 Future work
The implementation of the proposed conceptual framework will include an exploration
of different solutions for the user layer, to identify the best interaction mechanisms that
may sustain participation and collaboration to rule creation. In particular, we have al-
ready started such exploration by considering a variety of user-oriented tools in the AmI
and Internet of Things domains. The idea is to take inspiration for the design of an
interaction metaphor that is compatible with the interconnection metaphor of our con-
ceptual framework and able to address some challenges that the framework brings
about: rule definition and modification by different users, management of conflicting
rules, openness to environment and/or community extensions over time, simplicity and
expressiveness of single rules, overall readability of the configuration of the AmI envi-
ronment. In particular, the filtered lists metaphor of IFTTT and Atooma appear as the
most suitable for collaborative and distributed development of ECA rules, to be man-
aged by the underlying distributed inference engine foreseen in our framework. There-
fore, we will start with a comparison between these two systems. Another complemen-
tary line of research would pursue the perceived relative advantage of people already
using an AmI system since a while, to fill the gap in the literature between technical
feasibility and human-centred sustainability over time.
30
Proc. of Third International Workshop on Cultures of Participation in the Digital Age - CoPDA 2015
Madrid (Spain), May 26th, 2015 (published at http://ceur-ws.org).
Copyright © 2014 for the individual papers by the papers' authors. Copying permitted for private and academic purposes.
This volume is published and copyrighted by its editors.
References
1. Dahl, Y., Svendsen, R.-M.: End-User Composition Interfaces for Smart Environments: A
Preliminary Study of Usability Factors. In: Marcus, A. (ed.) Design, User Experience, and
Usability. Theory, Methods, Tools and Practice. LNCS, vol. 6770, pp. 118-127. Springer,
Berlin Heidelberg (2011)
2. Cabitza, F., Fogli, D., Piccinno, A.: Fostering Participation and Co-Evolution in Sentient
Multimedia Systems. In: 20th International Conference on Distributed Multimedia Systems
(DMS), pp. 47-56. Knowledge Systems Institute, Skokie, IL 60076, USA (2014)
3. Lieberman, H., Paternò, F., Wulf, V. (eds.): End User Development. Springer, Dordrecht,
The Netherlands (2006)
4. Sadri, F.: Ambient intelligence: A survey. ACM Computing Surveys 43(4), 1-66 (2011)
5. Danado, J., Paternò, F.: Puzzle: A mobile application development environment using a
jigsaw metaphor. Journal of Visual Languages & Computing 25(4), 297-315 (2014)
6. Demeure, A., Caffiau, S., Coutaz, J.: Activity based End-User-Development for Smart
Homes: Relevance and Challenges. Vol. 18. IOS Press (2014) 141 - 152
7. Davidoff, S., Lee, M., Yiu, C., Zimmerman, J., Dey, A.: Principles of Smart Home Control.
In: Dourish, P., Friday, A. (eds.) UbiComp 2006: Ubiquitous Computing. LNCS, vol. 4206,
pp. 19-34. Springer, Berlin Heidelberg (2006)
8. Cabitza, F., Gesso, I.: Reporting a User Study on a Visual Editor to Compose Rules in Active
Documents. In: Blashki, K., Isaias, P. (eds.) Emerging Research and Trends in Interactivity
and the Human-Computer Interface, pp. 182-203. IGI Global, Hershey, PA, USA (2014)
9. Ur, B., McManus, E., Pak Yong Ho, M., Littman, M.L.: Practical trigger-action
programming in the smart home. In: SIGCHI Conference on Human Factors in Computing
Systems, pp. 803-812. ACM, New York, NY, USA (2014)
10. García-Herranz, M., Haya, P., Alamán, X.: Towards a Ubiquitous End-User Programming
System for Smart Spaces. Journal of Universal Computer Science 16(12), 1633--1649
(2010)
11. Mavrommati, I., Darzentas, J.: End User Tools for Ambient Intelligence Environments: An
Overview. In: Jacko, J.A. (ed.) Human-Computer Interaction. Interaction Platforms and
Techniques. LNCS, vol. 4551, pp. 864-872. Springer, Berlin Heidelberg (2007)
12. Grasso, A., Convertino, G.: Collective Intelligence in Organizations: Tools and Studies.
Computer Supported Cooperative Work (CSCW) 21(4-5), 357-369 (2012)
13. Sadri, F., Stathis, K.: Ambient Intelligence. In: Rabuñal Dopico, J.R., Dorado, J., Pazos, A.
(eds.) Encyclopedia of Artificial Intelligence, pp. 85-91. IGI Global, Hershey, PA, USA
(2009)
14. Litvinova, E., Vuorimaa, P.: Engaging end users in real smart space programming. In: ACM
Conference on Ubiquitous Computing (UbiComp), pp. 1090-1095. ACM, New York, NY,
USA (2012)
15. Dey, A.K., Sohn, T., Streng, S., Kodama, J.: iCAP: Interactive Prototyping of Context-
Aware Applications. In: Fishkin, K.P., Schiele, B., Nixon, P., Quigley, A. (eds.) Pervasive
Computing. LNCS, vol. 3968, pp. 254-271. Springer, Berlin Heidelberg (2006)
31
Proc. of Third International Workshop on Cultures of Participation in the Digital Age - CoPDA 2015
Madrid (Spain), May 26th, 2015 (published at http://ceur-ws.org).
Copyright © 2014 for the individual papers by the papers' authors. Copying permitted for private and academic purposes.
This volume is published and copyrighted by its editors.