Ebook: Intelligent Environments 2009

As computers are increasingly embedded into our everyday environments, the objects therein become augmented with sensors, actuators, processors, memory, communication modules and novel interfaces. The capability for objects to perceive the environment, store and process data, pursue goals, reason about their intentions and coordinate actions in a holistic manner gives rise to the so-called Intelligent Environment (IE).

In such environments, real space becomes augmented with digital content, thus transcending the limits of nature and of human perception. The term “digital territory” (DT) was initially used, based on a real world metaphor: a territory is usually a continuum in space and has a measurable quantity of elements that are contained within its borders (only that the elements of a DT are active, as opposed to the usually passive objects found in real world territories). IEs constitute the incarnation of the Ambient Intelligence (AmI) vision within a DT. In the AmI paradigm, intelligent computation will be invisibly embedded into our everyday environments through a pervasive transparent infrastructure capable of recognizing, responding and adapting to individuals in a seamless and unobtrusive way.

IEs provide the necessary infrastructure and the services required by ubiquitous computing systems and applications, which are developed in order to support the realization of people's everyday tasks; in that sense, IEs host all technology mediated human activity. A “Virtual Residence” (the evolution of the physical home into the smart, digital home) was amongst the first IEs to be researched; the concept uses most of the notions that apply to “physical” residence, such as borders, markers, activities, etc and expands them to the digital space. The success of the IE conference series demonstrates the increasing popularity of IE research and systems, as the AmI vision continues to infuse all aspects of human life and activity, including architecture of private and public buildings, design of ecologies of smart objects and robots, health care, education, leisure and entertainment, even art and culture.

The 5th International Conference on Intelligent Environments (IE'09), held at the Politechnical University of Catalonia, Castelldefels, Barcelona, Spain, provides a collaborative forum for scientists, researchers and engineers from both industry and academia to present theoretical and practical results of Intelligent Environments research and their application. It is the next edition in a series of highly successful conferences that were organized in Colchester (UK), Athens (Greece), Ulm (Germany) and Seattle (USA) in the past four years. The conference programme includes full papers presented at special sessions and short papers from the doctoral colloquium and poster session. In addition, three thought provoking invited lectures are presented by Marcus Brunner, Carles Gomez Montenegro and Brian Johnson, on topics of current research, infused with the vision of Next Generation IEs. Collectively, this unique mix of interdisciplinary and visionary topics combines to provide another high quality IE conference.

However, none of these would have been possible without the team of distinguished colleagues and fellows who contributed their effort to making IE'09 possible. We cordially thank them all, starting from the local chairs, Dolors Royo, Angélica Reyes and Leandro Navarro and the publications chair, Michael Weber. Then there are the members of IE scientific committee, the session chairs and all the people serving in the Programme Committee, who worked hard publicizing the conference and collecting contributions. Finally, we extend our gratitude to our technical assistants, Kostas Togias, who managed the IE conferences web site, and Stefan Dietzel and Florian Schaub, who prepared the proceedings that you are now holding. But most of all, we'd like to thank you, the IE attendees, who have been supporting the IE conference during these years, contributing to making Intelligent Environments a reality that will, one day, embrace all activities of life.

Victor Callaghan and Achilles Kameas

IE'09 Programme chairs

This paper is meant to be an introduction both to the short story “Nebulous Mechanisms” as well as the concept of science fiction prototypes. In the tradition of the legendary Isaac Asimov, we jump into the long and symbiotic history of science fact and science fiction. Briskly we move through current work in the field and delve into the particulars of the craft of combining current scientific research, theory and practice with the fictional construct of the science fiction story. Finally we look into the specifics of the story “Nebulous Mechanisms” and conclude with the challenge to employ the science fiction prototype to foster innovation and explore the boundaries of modern science.

Wireless sensor networks (WSNs) have gained momentum as a technology for monitoring and control applications in a plethora of scenarios. Some examples include home automation, industrial control and environmental monitoring. In fact, these networks may enable the futuristic paradigm called ambient intelligence. Despite the challenges that present these networks, a variety of solutions (many of which are proprietary) have emerged. In parallel, the Internet Engineering Task Force (IETF) is devoting a lot of effort to the development of standard networking solutions for WSNs based on the Internet Protocol (IP). This work will allow straightforward interoperability between WSNs and the Internet and will also constitute a significant step towards the ‘Internet of Things’.

We note the disadvantages which many people with disabilities have regarding learning and using mathematics. We then describe work in progress on developing intelligent systems to assist such people by allowing them to dictate mathematical expressions in relatively natural language into either a system which can interpret and display the expressions in a document or web page, or a web-based calculator, including the possibilities of feedback in both visible and other modes (e.g. synthesized speech and Braille). We discuss the implementation and use of these systems, initial results of a study to evaluate their utility, and potential further developments.

This paper presents a speaker classification architecture for VoiceXML-based applications. Our analysis component receives user utterances from the VoiceXML platform, performs feature extraction and classifies speaker characteristics such as age, gender and emotional state. This additional information about the speaker can be employed to adapt system prompts and the dialogue strategy to specific user groups. The implementation of our prototype shows, that speaker classification is feasible without significant delays and with high accuracies of over 95% for anger detection and gender classification.

This paper proposes a Model-View-Controller architecture for Spoken Dialogue Management (SDM) that consists of VoiceXML as View, i.e., as presentation layer and a Controller layer using a Model to determine the next possible system utterance. In this paper we focus on the Model layer and present a Dialogue Domain Ontology, which combines the Information State and the ontology theory. The Dialogue ontology can be extended by several optional components such as a domain reasoner, which supports and governs the ontology, an analysis layer to classify the user or a translation layer, which triggers various applications. Our work aims at a unified model for SDM related information handling for various tasks in intelligent environments.

Recent research shows that including logical reasoning into dialogue management could be of enormous benefit for the capabilities of the system. However logical reasoning is to some extent limited in its possibilities to infer or to find a conclusion given certain premises. Furthermore reasoning systems heavily rely on backtracking algorithms which is a great shortcoming concerning performance aspects.

In order to improve the performance as well as the reasoning capabilities of the dialogue manager we propose an architecture that is based on logical reasoning but uses a constraint solving mechanism for enlarging the logical capabilities of the dialogue manager.

This work describes a multimodal product information and comparison system designed for a rural corner store (the so called Dorfladen). Such a Dorfladen only provides a limited assortment, which we extend with virtual items on embedded displays, allowing customers to inspect and select unavailable products along their real world shopping activities for later ordering. We present the design and implementation of the system prototype and discuss the main results of a formal user study evaluating the user interface, which combines tangible interaction and natural language dialogues to interact with real and virtual products in one instrumented space.

Social connectedness, i.e. the experience of belonging and relatedness between people, is a central concept in understanding and evaluating communication media, in particular awareness systems. A generic measure based on this construct an support the design of such systems. The current paper describes the construction of two questionnaires for the measurement of this concept. These questionnaires were subsequently applied in survey studies in order to establish the structure of the concept and to identify the items that are suited for the measurement of its dimensions. One questionnaire was subjected to an initial validation. We conclude with some preliminary suggestions regarding (design) approaches to foster social connectedness.

In this paper, a novel infrastructure, called Medient is proposed. The name “Medient” is constructed from “ambient” and “media”. Medient is a service oriented architecture using web services to create context-aware multimedia services on mobile devices. We outline the technical approach using Bluetooth device detection and reverse AJAX push service creation. Results, usage showcases and a demo scenario are given for the Holstentor museum in Luebeck, Germany, as a typical example of a dense information environment with an increasing demand for contextualized exhibition services.

This paper presents possibilities offered by public intelligent environment to foster learning through serendipitous interaction among uncoupled users. Four possible situations of how serendipitous interaction may occur and lead to learning activities are presented. In these situations capturing the right information, modeling it and providing awareness information back to the users plays a major role. Therefore elements that make up this awareness information are presented and a model for fostering serendipitous interaction in public intelligent environments is also proposed.

Studies in economic psychology have repeatedly revealed that energy consumption behavior in the household could be optimized through the provision of frequent and immediate feedback. This aspect is still widely neglected by conventional metering equipment installed in most homes. Energy meters are generally placed at quite hidden locations and reading them only once a year or at best once a month for billing purposes remains the rule. Especially observation of variations or trends over longer periods of time is cumbersome and has to be done on the basis of handwritings or transferring manually read data to applications like spreadsheets. We introduce Perceivable Energy, a system that supports pervasive approaches to raise awareness of energy consumption. More innovative feedback mechanisms like, e.g., ambient displays are backed by a more conventional graphical user interface to provide detailed consumption data on demand. The system is designed to make integration and interconnections of new metering devices and visualizations simple and constitutes a good basis for further research on awareness of energy usage in our (future) living environments.

This paper discusses the requirements of future Home Area Networks (HAN) with respect to resource discovery. We discuss existing methods and provide some experimental data that shows them to be unsuitable for future HANs. In response, we present a new protocol called “Entity Resolution Protocol” (ERP) and experimental data which shows that it significantly outperforms the other methods discussed.

A knowledge representation system is introduced that allows the recognition of temporal patterns of events in context-aware environments. The system is based on standard frameworks, such as an ontology, an inference mechanism, and relational operators acting on numerical quantities. The paper describes how knowledge is managed, then introduces a collection of temporal operators that are inspired by the Allen's interval algebra, and then details a situation recognition algorithm to assess knowledge semantics. An example is reported to describe the approach.

Intelligent environments are increasingly rich in ubiquitous sensing capabilities that can be leveraged to know which actions a user is engaged in at any given moment in time. The ability of an intelligent environment to recognize a high-level plan of activities performed by the user in a smart home would allow to construct proactive services, such as reminding, forecasting and providing timely physical support. This article proposes an approach to human activity recognition based on temporal planning. The approach leverages on one hand the ubiquitous sensors provided by the PEIS-Home, a sensor-rich intelligent environment, and, on the other hand, the temporal representation and reasoning capabilities of OMPS, a constraint-based temporal planning and scheduling framework.

Does chance have a role in intelligent environments? In this work-in-progress paper we argue that chance and non-deterministic behaviour can play a fundamental and important role in intelligent environments. We discuss how this behaviour can be both destructive and constructive. Underpinning our ideas is the view that intelligent environments may be seen as a complex system of interacting services. In the first part of this paper we show that such complex systems can produce unexpected interactions that cause unplanned and often undesirable instabilities. However, not all instabilities are undesirable and in the second half of this paper, we present a conceptual notion that views system instability as a form of irrationality and propose a quantum control model for service agents within smart environments. We conjecture that irrational control models enable the service agents to perform better than if they were using traditional, rational, control models. Our purpose in presenting this work is to both provoke discussion and describe our early research on what we hope will be an interesting direction for intelligent building research.

Little is reported in the literature on the methodology used to develop Intelligent Environments. These are inherently complex systems as they are the product of a multidisciplinary approach. Their main purpose is to help people directly, which leads to these environments to take on enormous responsibilities of safety and well being of the users of the environment. We illustrate in this paper how techniques and tools developed in other areas of Computer Science to increase the reliability of software can be applied at design and modelling stages to guide the engineering and developing team to reduce the potential for future malfunctioning.

Smart environments rely on artificial intelligence techniques to make sense of the sensor data and to use the information for recognition and tracking activities. However, many of the techniques that have been developed are designed for simplified situations. In this paper we discuss a more complex situation, namely recognizing activities when they are interweaved in complex and realistic scenarios. This technology is beneficial for monitoring the health of smart environment residents and for correlating activities with parameters such as energy usage. We describe our approach to interleaved activity recognition and evaluate various probabilistic techniques for activity recognition. We validate our algorithm on real sensor data collecting in our smart apartment testbed.

Intelligent Environments depend on their capability to understand and anticipate user's habits and needs. Therefore, learning user's common behaviours becomes an important step towards allowing an environment to provide such personalized services. We have developed a system which learns user's patterns taking into account the special features of Intelligent Environments. Due to the complexity of the system, this paper will focus on a specific part of such a system, the module that discovers what sets of actions are frequently carried out by the user.