La rencontre de l’intelligence ambiante et de la robotique d’assistance a donné naissance à ce que nous avons appelé robotique ambiante d’assistance car destinée à assister une personne en perte d’autonomie. Elle s’appuie sur l’existence de réseaux d’objets communicants présents dans l’environnement de la personne pour décliner un ensemble de services et de téléservices destinés à faciliter la vie quotidienne de cette personne et de son entourage. Un, voir plusieurs robots peuvent être présents dans cet environnement. Une communauté scientifique récente s’est construite autour de la robotique ubiquitaire. Tandis que les robots des générations précédentes ont été conçus pour réaliser des tâches spécifiques et construits en tant qu’unité indépendante, la nouvelle génération vise l’ubiquité. L’autonomie du robot est obtenue par une interaction étroite entre le robot et l’environnement ambiant communicant. Dans le contexte de l’intelligence ambiante, les objets communicants de l’environnement peuvent jouer un rôle "facilitateur" en aidant le robot à se localiser, naviguer, rechercher un objet. Inversement, le robot peut être vu comme un objet communicant qui est mis à contribution par des services autres que l’assistance à la personne en perte d’autonomie. Ainsi le premier objectif de cette thèse est de proposer une architecture informatique permettant la coopération entre le robot et les objets communicants présents dans l’environnement tels que les capteurs ou les actionneurs. Le deuxième objectif est de proposer une méthode de localisation par coopération robot-environnement communicant. Notre ambition est de localiser avec certitude le robot dans une zone de l’habitat. / The meeting of ambient intelligence and assistive robotics gave birth to ambient assistive robotics which aims at assisting a person in loss of autonomy. It relies on the existence network of communicating objects present in the environment of the person. The objective is to propose a set of services and teleservices for facilitating the person's daily life. One or more robots can be present in this environment. Previous generation of robots have been designed to perform tasks specific and built as an independent unit, while the new generation of robots tries to take advantage of ubiquity. In this framework, the robot autonomy is obtained by a close interaction between the robot and the ambient environment. Up to recent years the robot moved in a rather hostile environment which did not facilitate its task. In the context of ambient intelligence, the communicating objects of ambient environment can play a "facilitator" role in helping the assistive robot for locating itself, searching for a person … Conversely, the robot can be seen as a communicating object which is used by services to the person in loss of autonomy. The first objective of this thesis is to propose an architecture for cooperation between the robot and communicating objects and the second objective is to propose a localization method by cooperation of ambient environment and a mobile robot.

This thesis explores the trajectory that the developing technological fields of Ambient Intelligence and Persuasive Technologies introduce new intricate relationships beyond fundamental use and availability because they change our abilities to act. Since its classic articulation by Hegel (1927) philosophical explication of the relationship between people and technology states that technology is a mediating factor between people and the world. Associated with this view, which has characterized the resulting phenomenology and philosophy of technology for nearly two decades, is an understanding of technology as a form of alienation. In this dissertation the author shows how this old interpretation of the relationship between a person and their tool has emphasized how the person is active whilst the tool is passive. This traditional distinction fails to grasp the complex interaction between people and technology in the contemporary world. The nature of new technologies and novel theoretical work in this field suggests that this critical framework is now inadequate. Today, technology mediates the relationship between people and the world in increasingly complex and often collective ways. McLuhan (1967) stated: “Media evoke in us unique ratios of sense perceptions. The extension of any one of these senses alters the way we think and act”. As Greenfield (2006) and Fogg (2002) also posit, certain Ambient Intelligence and Persuasive Technologies are in-principle shaping everyday human behaviours in radically new ways. In particular, I explore how new technologies like those developed in the Artificial Companions Project can impact on our understanding of intimacy and identity. Indeed, Ambient Intelligence Technologies may play the role of reference groups (Shibutani 1987), groups who are real or imaginary and whose standpoints are being used as the frame of reference for the human actor. Given that these technologies have continuously reconfigured identification and profiling practices, this analysis rephrases insight of philosophers like Paul Ricoeur (1990), George Herbert Mead (1959) and Helmuth Plessner (1975) to trace how: The construction of our identity is mediated by how we profile others as profiling us. Thus, new technologies can become reference groups, encroaching on our everyday activities and even affecting our moral decision-making processes. As genuine upgrades of our practical space, they are destined to play a larger formative role in people’s lives in the future. Following Heidegger in Das Ding (Heidegger 1951), Latour once framed the wider social role of technologies as res publica or ‘public things’ (Latour 2005). He pointed out that the old German word ‘ding’ etymologically did not only infer ‘material object’ but also to assembly as gathering space - that thing that can bring together what it separates. Following Latour, Verbeek states that technological ‘things’ do not only mediate our existence, but are places where these mediations are made explicit – therefore, Verbeek argues, they are the places where people have to start to discuss and criticise the quality of the ways in which these ‘things’ help to shape our daily lives (Verbeek 2008). This thesis attempts to offer a new approach to this criticism through theoretical comparison and transdisciplinary analysis.

L'informatique sensible au contexte vise à réduire la quantité d'informations explicites qu'un utilisateur doit fournir pour que le système accomplisse la tâche souhaitée. Ceci est particulièrement vrai dans le domaine récent de l'intelligence ambiante où les objets de la vie courante deviennent capables de déclencher une action ou un échange spontané d'informations, sans interaction avec l'utilisateur. Les progrès techniques en matière de réseaux de communication sans fil, d'équipements mobiles individuels, de capteurs et de logiciels embarqués, rendent aujourd'hui possibles des services aux usagers dépendants du contexte, mais les applications concrètes demeurent encore très limitées. Les travaux existants dans la littérature décomposent la gestion de contexte en quatre fonctionnalités: la collecte, l'interprétation, la détection de situations et l'utilisation pour l'adaptation. L'élément discriminant des solutions existantes est la qualité des informations abstraites obtenues par inférence et devant caractériser les situations de l'utilisateur. Les limites de ces solutions sont le manque de composition aisée des informations de contexte, le passage à l'échelle, tant en termes de quantité d'informations de contexte que de nombre d'applications clientes, l'absence de garantie sur la cohérence et la qualité des informations de contexte, et le manque de solutions intergicielles permettant de libérer le concepteur d'applications des aspects liés à la gestion de contexte. Nous nous intéressons dans cette thèse à la gestion de la qualité de contexte (QoC) dans un environnement ambiant. Les problématiques de gestion de la qualité de contexte sont multiples: choisir la méthode adéquate pour la gestion du contexte, extraire la qualité associée au contexte, interpréter et analyser la qualité de contexte pour les applications sensibles au contexte. Nous proposons de répondre à ces problématiques en intégrant la qualité de contexte au sein de la plateforme de gestion de contexte COSMOS (http://picoforge.lntevrv.fr/proiects/svn/cosmos) de l'équipe MARGE (http://www-inf.itsudparis.eu/MARGE) de Télécom SudParis. Afin d'effectuer cette intégration, nous avons conçu des éléments spécifiques à la qualité de contexte et avons mis en place une gestion fine et efficiente de cette qualité en limitant le surcoût associé. Nous proposons également un processus de conception basé sur l'ingénierie dirigée par les modèles afin de générer les éléments requis à la gestion de la qualité de contexte. Nous avons validé nos contributions à l'aide de deux applications fonctionnant sur téléphone mobile : une application de "vente flash" dans un centre commercial et une application de détection de localisation sur un campus. Les tests de performances que nous avons effectués permettent de comparer les résultats avec et sans la prise en compte de la QoC et montrent le faible coût de la gestion de la qualité par rapport aux améliorations apportées aux applications sensibles au contexte / Context-aware computing aims to reduce the amount of explicit information required from a user for a system to perform a task. This is particularly true in the recent domain of ambient intelligence where everyday life objects are able to trigger an action or a spontaneous information exchange, without any interaction with the user. Technical advances in wireless communication, personal mobile devices, sensors and embedded software make context-aware services possible, but concrete applications are still very limited. The solutions proposed in the literature decompose context management into four functions: acquisition, interpretation, situation detection and application adaptation. The differentiating element in these proposals is the quality of the high-level context information obtained by inference and characterising the situation of the user. The limits of these solutions are the difficulty for composing context information scalability in terms of the quantity of context information and of the number of client applications, the absence of guarantee on the consistency of context information and the lack of middleware solutions able to free the designer of context-aware applications from the management of context data. In this thesis, we are interested in the management of the quality of context information (QoC) in an ambient environment. There are several key issues in QoC management: choosing the adequate method for context management, extracting the quality associated to the context, analysing and interpreting the quality of the context with regard to the requirements of context-aware applications. We propose to answer these questions by integrating QoC management into the COSMOS context management framework (http://picoforge.int-evry.fr/projects/svn/cosmos) developed by the MARGE team (http://www-inf.itsudparis.eu/MARGE) of Télécom SudParis.For this purpose, we have designed the necessary components dedicated to QoC management and we have implemented the mechanisms allowing a fine-grain manipulation of the QoC together with a limitation of the associated overhead. We also propose a design process based on model-driven engineering in order to automatically generate the elements responsible of QoC management. We validate our contributions through the development of two prototype applications running on mobile phones: a Flash sale offer application to be used in malls and a location detection application proposed to the students of a campus. The performance tests we have conducted allow to compare the results obtained with and without taking into account the QoC and show the low overhead associated to QoC manaqement with regard to the benefits brought to context-aware applications and services

<p>This report presents the existence of a set of problems making the growth of the domotic field more difficult. They are mainly the lack of a common communication standard among devices and the existence of a proprietary market, where each provider focuses on developing its own devices, protocols and interfaces.</p><p>There isn’t a convergence criterion in order to overcome this problem by the main domotic providers. Several studies try to overcome this problem by applying different strategies. This study analyses the main strategies followed in that field, concluding with a model that combines them. The model is based on the use of OPC and web services.</p>

- / -

This report presents the existence of a set of problems making the growth of the domotic field more difficult. They are mainly the lack of a common communication standard among devices and the existence of a proprietary market, where each provider focuses on developing its own devices, protocols and interfaces. There isn’t a convergence criterion in order to overcome this problem by the main domotic providers. Several studies try to overcome this problem by applying different strategies. This study analyses the main strategies followed in that field, concluding with a model that combines them. The model is based on the use of OPC and web services.

L'objectif de ce travail de thèse est de proposer un canevas sémantique intégrant un méta-modèle et des outils de raisonnement permettant à tout concepteur de système ubiquitaire de mettre en oeuvre facilement des mécanismes de gestion des politiques de la vie privée. Le canevas proposé intègre une architecture middleware générique qui offre des composants pour définir, administrer et contrôler l'application des politiques de confidentialité. Notre approche proposée est hybride. Elle est fondée sur l’ingénierie dirigée par les modèles et sur un raisonnement à base d'ontologies et de règles d'inférence opérant selon l'hypothèse du monde clos. Le méta-modèle proposé est caractérisé par un niveau d'abstraction et d'expressivité élevé permettant de définir des politiques de gestion de la vie privée indépendamment du domaine d'application pouvant être adaptées à différents contextes. Il définit, aussi, un cadre conceptuel pour établir des modèles de règles génériques et décidables permettant de prendre des décisions de contrôle cohérentes pour la protection de la vie privée. Ces modèles de règles sont mis en oeuvre grâce au langage de règles SmartRules permettant de mettre en oeuvre un contrôle adaptatif. Ce dernier est basé sur un raisonnement non-monotone et une représentation des instances de concepts selon la supposition du nom unique. Nous avons validé le canevas proposé à travers un scénario typique mettant en oeuvre des services d'assistance ambiante sensibles à la vie privée de personne âgée. / This thesis aims at proposing a semantic framework that integrates a meta-model and reasoning tools allowing any ubiquitous system designer to easily implement mechanisms to manage privacy policies. The proposed framework includes a generic middleware architecture that provides components to define, manage and monitor the implementation of privacy policies. Our approach is an hybrid one based on Model-Driven Engineering and a reasoning based on ontologies and inference rules operating on the assumption of the closed world. The proposed meta-model is characterized by a high level of abstraction and expressiveness to define privacy policies management regardless of the domain application and can be adapted to different contexts. It defines, also, a conceptual framework for generic decidable modelling rules to make consistent control decisions on user privacy. These model rules are implemented using the SmartRules language that could implement an adaptive control. The latter is based on a non-monotonic reasoning and representation of instances of concepts according to the unique name assumption. We have validated the proposed semantic framework through a typical scenario that implements support ambient intelligence privacy-aware services for elderly.

<p>This work focuses on the socio-technical aspects of artificial intelligence, namely how (specific types of) intelligent systems function in human workplace environments. The goal is first to get a better understanding of human needs and expectations when it comes to interaction with intelligent systems, and then to make use of the understanding gained in the process of designing and implementing such systems.</p><p>The work presented focusses on a specific problem in developing intelligent systems, namely how the artefacts to be developed can fit smoothly into existing socio-cultural settings. To achieve this, we make use of theories from the fields of organisational psychology, sociology, and linguistics. This is in line with approaches commonly found in AI. However, most of the existing work deals with individual aspects, like how to mimic the behaviour or emulate methods of reasoning found in humans, whereas our work centers around the social aspect. Therefore, we base our work on theories that have not yet gained much attention in intelligent systems design. To be able to make them fruitful for intelligent systems research and development, we have to adapt them to the specific settings, and we have to transform them to suit the practical problems at hand.</p><p>The specific theoretical frameworks we draw on are first and foremost activity theory and to a lesser degree semiotics. Activity theory builds on the works of Leont'ev. It is a descriptive tool to help understand the unity of consciousness and activity. Its focus lies on individual and collective work practise. One of its strengths, and the primary reason for its value in AI development, is the ability to identify the role of material artefacts in the work process. Halliday's systemic functional theory of language (SFL) is a social semiotic theory that sets out from the assumption that humans are social beings that are inclined to interact and that this interaction is inherently multimodal. We interact not just with each other, but with our own constructions and with our natural world. These are all different forms of interaction, but they are all sign processes.</p><p>Due to the obvious time and spatial constraints, we cannot address all of the challenges that we face when building intelligent artefacts. In reducing the scope of the thesis, we have focused on the problem of explanation, and here in particular the problem of explanation from a user perspective. By putting social theories to work in the field of artificial intelligence, we show that results from other fields can be beneficial in understanding what explanatory capabilities are needed for a given intelligent system, and to ascertain in which situations an explanation should be delivered. Besides lessons learned in knowledge based system development, the most important input comes from activity theory.</p><p>The second focus is the challenge of contextualisation. Here we show that work in other scientific fields can be put to use in the development of context aware or ambient intelligent systems. Again, we draw on results from activity theory and combine this with insights from semiotics.</p><p>Explanations are themselves contextual, so the third challenge is to explore the space spanned by the two dimensions ability to explain and contextualisation. Again, activity theory is beneficial in resolving this issue.</p><p>The different theoretical considerations have also led to some practical approaches. Working with activity theory helps to better understand what the relevant contextual aspects of a given application are and helps to develop models of context which are both grounded in the tradition of context aware systems design and are plausible from a cognitive point of view.</p><p>Insights from an analysis of research in the knowledge based system area and activity theory have further lead to the amendment of a toolbox for requirements engineering, so called problem frames. New problem frames that target explanation aware ambient intelligent systems are presented. This is supplemented with work looking at the design of an actual system after the requirements have been elicited and specified. Thus, the socio-technical perspective on explanations is coupled with work that addresses knowledge representation issues, namely how to model sufficient knowledge to be able to deliver explanations.</p>

This work focuses on the socio-technical aspects of artificial intelligence, namely how (specific types of) intelligent systems function in human workplace environments. The goal is first to get a better understanding of human needs and expectations when it comes to interaction with intelligent systems, and then to make use of the understanding gained in the process of designing and implementing such systems. The work presented focusses on a specific problem in developing intelligent systems, namely how the artefacts to be developed can fit smoothly into existing socio-cultural settings. To achieve this, we make use of theories from the fields of organisational psychology, sociology, and linguistics. This is in line with approaches commonly found in AI. However, most of the existing work deals with individual aspects, like how to mimic the behaviour or emulate methods of reasoning found in humans, whereas our work centers around the social aspect. Therefore, we base our work on theories that have not yet gained much attention in intelligent systems design. To be able to make them fruitful for intelligent systems research and development, we have to adapt them to the specific settings, and we have to transform them to suit the practical problems at hand. The specific theoretical frameworks we draw on are first and foremost activity theory and to a lesser degree semiotics. Activity theory builds on the works of Leont'ev. It is a descriptive tool to help understand the unity of consciousness and activity. Its focus lies on individual and collective work practise. One of its strengths, and the primary reason for its value in AI development, is the ability to identify the role of material artefacts in the work process. Halliday's systemic functional theory of language (SFL) is a social semiotic theory that sets out from the assumption that humans are social beings that are inclined to interact and that this interaction is inherently multimodal. We interact not just with each other, but with our own constructions and with our natural world. These are all different forms of interaction, but they are all sign processes. Due to the obvious time and spatial constraints, we cannot address all of the challenges that we face when building intelligent artefacts. In reducing the scope of the thesis, we have focused on the problem of explanation, and here in particular the problem of explanation from a user perspective. By putting social theories to work in the field of artificial intelligence, we show that results from other fields can be beneficial in understanding what explanatory capabilities are needed for a given intelligent system, and to ascertain in which situations an explanation should be delivered. Besides lessons learned in knowledge based system development, the most important input comes from activity theory. The second focus is the challenge of contextualisation. Here we show that work in other scientific fields can be put to use in the development of context aware or ambient intelligent systems. Again, we draw on results from activity theory and combine this with insights from semiotics. Explanations are themselves contextual, so the third challenge is to explore the space spanned by the two dimensions ability to explain and contextualisation. Again, activity theory is beneficial in resolving this issue. The different theoretical considerations have also led to some practical approaches. Working with activity theory helps to better understand what the relevant contextual aspects of a given application are and helps to develop models of context which are both grounded in the tradition of context aware systems design and are plausible from a cognitive point of view. Insights from an analysis of research in the knowledge based system area and activity theory have further lead to the amendment of a toolbox for requirements engineering, so called problem frames. New problem frames that target explanation aware ambient intelligent systems are presented. This is supplemented with work looking at the design of an actual system after the requirements have been elicited and specified. Thus, the socio-technical perspective on explanations is coupled with work that addresses knowledge representation issues, namely how to model sufficient knowledge to be able to deliver explanations.

Zugl.: Braunschweig, Techn. University, Diss., 2009.