隨著ontology應用及語意網的普及，可以預見將有越來越多的應用系統以ontology做為異質系統間之資料交換格式。若現行的分散式監控系統，也能夠以ontology做為程式外部的溝通及資料描述介面，那不同的監控系統彼此要交換資料就會更容易。此外，系統本身若也採用ontology做為與外部溝通的語言，則系統在面對裝置異動或邏輯修改時，將能更快速的反應需求也能更有彈性的面對異動。本研究的目標，便是想要從監控系統中分離出程式應用邏輯及設備描述邏輯，改以ontology的描述方式來取代，進而建立一套以ontology為核心的監控系統，並藉由ontology的導入來改善現行監控系統修改不易及彈性不大的缺點。 為了驗證上面的論述，我們設計了一套以ontology為組態描述的webpanel系統。製造商可以用符合ontology的語法來描述設備的規格，系統整合商也能透過ontology設計監控系統的應用邏輯，結合兩者的ontology描述，便能讓webpanel系統執行監控程序。透過我們的實驗證明，以ontology做為監控系統的應用邏輯描述是可行的。在建構監控系統的過程中，系統整合商能很快的藉由應用系統描述檔完成監控系統的描述，同時webpanel在面對設備故障發生時，也能迅速的透過ontology的搜尋，找到此區域中可用的設備來替代。 對系統整合商來說，建置另一套全新的監控系統只需依據新環境的狀況，適度的調整應用系統描述檔，便可以開始進行監控系統安裝及測試，比起傳統每換一個環境就需要重新撰寫程式的作法會快上許多。另外，藉由ontology描述的特性，可以讓設計的經驗完整的保存下來，供重複使用，對於人才流動快速的企業來說，也可以有效的降低訓練人力及開發系統的時間。 我們希望能夠藉由設計webpanel過程中獲得的相關經驗來簡化傳統監控系統的設計流程及維護時間，將ontology的優點導入相對封閉的監控領域中，讓其更易與其他領域相結合，為人類生活帶來更大的便利。 / As ontology becomes more common and more popular, we are expecting to see an increase in its usage among heterogeneous applications with a need of information exchange. If current distributed monitoring system can utilize ontology as an external communication platform, it will be much easier for different monitoring systems to share and exchange information. In addition, if the systems are able to use ontology as its external communication language, it will be more flexible and faster for the system to make necessary logical changes. The purpose of the research is to differentiate application logic and description logic out of a monitoring system, replace them with ontology, and eventually make it the core of the monitoring system. This could overcome the drawbacks of the current monitoring systems such as lack of flexibility. In order to justify the objective, we designed a system called webpanel . Manufacturers can use languages that are compatible to ontology to describe the equipment’s specification. System integrators can also use ontology to design system application logic. With these two types of information, we can allow the webpanel to conduct process monitoring. Through our research we found that using ontology to monitor application logic is a good way to construct a flexible surveillance system. During the construction period, system integrators can finish the monitoring description promptly based on the system description. When an equipment failure is encountered, though the search capability of ontology, the system can automatically look for replacement equipment within the area. From a system integrator’s perspective, rebuilding a brand new monitoring system only depends on some factors in the new environment. With a few minor changes in the system description file, we can deploy a monitoring system shortly. It is highly flexible compared to the traditional way where a new environment will need a new set of programs with customized application logic. Ontology allows us to preserve reusable design experience. Companies that are facing huge turn-over rates can really benefit from it by reducing the cost of training and system development. We are not the first one to use ontology but we hope that through the design of webpanel, we can acquire experiences to help simplify design in traditional monitoring system and maintenance time. We hope by introducing ontology into the conservative monitoring industry, it can help connect with other industries and bring greater convenience to people’s life.

近年來科技匯流發展日行千里，新興無線寬頻的接取技術（Broadband Wireless Access）不斷升級，為民眾的生活帶來更便利的創新應用，而寬頻接取服務的亦由以往一種技術提供一種服務的方式，逐漸轉換為涵蓋數據（Data）、語音（Voice）及影音內容（Video）等三種服務的匯流。加速了電信、廣電與網路通訊產業升級的需求，也帶來了新的市場動能與發展契機。 本研究以台灣為研究範本，以個案研究方法來探討在網路為基礎的數位匯流環境下數位生活之發展趨勢、個案公司科技產品未來的發展趨勢及因應策略。 而本研究發現未來家用閘道器產品市場，將會朝向整合家中所有可上網設備的整合性產品特性發展。亦將會是國內網通設備商值得切入的新領域。娛樂型家用閘道器，目前市場上多以機上盒(Set top box, STB)加上Router的功能型態出現，惟必須等待STB零售市場打開後，國內的廠商將會有機會創造更大利潤空間。而個案公司在研發方向應以Cable Docsis的核心技術為中心,從水平方向擴展至家庭網路的Home applications。追求更快的速度,更多元化的網路技術和更方便的寬頻服務。在縱向方面,我們則以完整的解決方案為主軸,往頭端CMTS技術和網管為發展重點。

Based on the fact that many electronic devices are digitalized in our world in order to facilitate our lives, there is a large potential for development in the home. Smart Home solutions are giving us the opportunity to control and manage for example alarms, electricity and surveillance but the technology's rapid improvement paves the way for issues related to security. The objectives for this work will bring up common communication technologies, security and vulnerabilities in the context of a Smart Home and what could be done for future work. In order to investigate the objectives, a literature study has been conducted together with an experiment. The experiment result exploits a weakness in a common Smart Home technology used in the network enable devices in the form of threats and vulnerabilities. In order to mitigate and minimize threats and vulnerabilities one conclusion is that a security policy could be produced. This security policy should provide the user with good practice of how to manage security in order to mitigate vulnerabilities and threats within the Smart Home.

Les progrès de la technologie des capteurs et leur disponibilité ont permis de mesurer diverses propriétés et activités des habitants dans une maison intelligente. Cependant, l’obtention de connaissances significatives à partir d’une grande quantité d’informations collectées à partir d’un réseau de capteurs n’est pas une tâche simple. En raison de la complexité du comportement des habitants, l’extraction d’informations significatives et la prédiction précise des valeurs représentant les activités futures d’un occupant sont des défis de recherche [6]. L’objectif principal de notre travail de thèse est d’assurer une analyse efficace des données recueillies à partir des capteurs d’occupation dans une maison intelligente. A ce propos, ce travail se base sur la reconnaissance et l’évaluation des activités quotidiennes d’une personne âgée dans le but d’observer, de prédire et de suivre l’évolution de son état de dépendance, de santé et de détecter par la même occasion, la présence d’une perte ou d’une perturbation de l’autonomie en temps réel. / Advances in sensor technology and their availability have measured various properties and activities of residents in a smart home. However, obtaining significant knowledge from a large amount of information collected from a sensor network is not a simple task. Due to the complexity of the behavior of the inhabitants, the extraction of meaningful information and the accurate prediction of values ​​representing the future activities of an occupant are research challenges. The main objective of our thesis work is to ensure an efficient analysis of data collected from occupancy sensors in a smart home. In this regard, this work is based on the recognition and evaluation of the daily activities of an elderly person in order to observe, predict and monitor the evolution of his state of dependence, health and to detect by the same occasion, the presence of a loss or a disruption of autonomy in real time.

Les maisons intelligentes sont des habitations équipées de technologies de l'information et de la communication qui anticipent et répondent aux besoins des occupants. Malgré les nombreux travaux et solutions existants, seulement peu d'exemplaires de maisons intelligentes ont été construits et vendus. La raison cachée derrière cette lente diffusion est l'orientation technologique des approches existantes. Cette thèse de doctorat vise à démontrer qu'une maison intelligente peut fournir des fonctionnalités conçues avec une approche centrée utilisateur, en prenant en compte de considérations ergonomiques sur l'activité domestique et sur la cognition humaine. Ceci est réalisé en collaboration avec des ergonomes cognitifs, qui aident à "prendre garde" à l'écart entre le contexte humain et le contexte compréhensible par une machine. En utilisant une instrumentation légère, qui minimise les problèmes d'acceptabilité et de protection de la vie privée, ce travail de thèse a mené aux contributions suivantes: (i) la conception interdisciplinaire de fonctionnalités adaptées, en collaboration avec des ergonomes cognitifs; (ii) la conception d'un système sensible au contexte qui capture et raisonne sur des informations contextuelles incertaines de façon distribuée; (iii) la réalisation d'un prototype qui démontre la fourniture de fonctionnalités qui réalisent des économies d'énergie, tout en préservant le confort des habitants. / Smart homes are residences equipped with information and communication technologies that anticipate and respond to the needs of the occupants. Despite the numerous research and industrial efforts, today only few expensive smart homes have been built and sold. The reason behind this slow uptake is the technology-driven approach characterizing existing solutions. The doctoral Thesis aims at demonstrating that a smart home can provide functionalities designed with a user-centered approach, taking into account ergonomic considerations about domestic activity and human cognition. This is achieved in collaboration with cognitive ergonomists, which help "minding the gap" between human context and machine-understandable context. Using off-the-shelf and lightweight instrumentation (also minimizing privacy concerns), extending existing context modeling, reasoning and management tools and following the Ubiquitous Computing principles, the doctoral work led to the following achievements: (i) the inter-disciplinary design of suitable functionalities, in collaboration with cognitive ergonomists; (ii) the design of a context-aware system that captures and reasons about uncertain contextual information in a distributed fashion; (ii) the realization of a working prototype that demonstrates the provision of energy-saving and comfort-preserving functionalities.

Energy consumption in the form of fuel or electricity is ubiquitous globally. Among energy types, electricity is crucial to human life in terms of cooking, warming and cooling of shelters, powering of electronic devices as well as commercial and industrial operations. Therefore, effective prediction of future electricity consumption cannot be underestimated. Notably, repeated imbalance is noticed between the demand and supply of electricity, and this is affected by different weather profiles such as temperature, wind speed, dew point, humidity and pressure of the electricity consumption locations. Effective planning is therefore needed to aid electricity distribution among consumers. Such effective planning is activated by the need to predict future electricity consumption within a short period and the effect of weather variables on the predictions. Although state-of-the-art techniques have been used for such predictions, they still require improvement for the purpose of reducing significant predictive errors when used for short-term load forecasting. This research develops and deploys a near-zero cooperative probabilistic scenario analysis and decision tree (PSA-DT) model to address the lacuna of significant predictive error faced by the state-of-the-art models and to analyse the effect of each weather profile on the cooperative model. The PSA-DT is a machine learning model based on a probabilistic technique in view of the uncertain nature of electricity consumption, complemented by a DT to reinforce the collaboration of the two techniques. Based on detailed experimental analytics on residential, commercial and industrial data loads, the PSA-DT model with weather profiles outperforms the state-of-the-art models in terms of accuracy to a minimal error rate. This implies that its deployment for electricity demand planning will be of great benefit to various smart-grid operators and homes. / School of Computing / M. Sc. (Computer Science)

This study is an attempt for providing a socio-historical perspective to the smart home concept that is a proposal for the future domestic sphere by the application of intelligent technologies. For achieving this attempt, the smart home is investigated with its roots within the social history of domestic technology, by posing the question of a relation between the machine house idea of modern architecture and the smart home concept as the main question of the research. After an inquiry on the smart home concept and on the future private sphere that is envisaged within it, the emergence of modern private sphere, the transformation of private sphere under the impacts of industrialization, the appearance and the social consequences of the machine house idea, the introduction of mass produced appliances to the home and the impacts of it on private sphere, the development of communication and information technologies and their domestication processes are investigated. The coconstructive relationship of technology with the private sphere is pursued throughout this investigation. The research questions of the study are answered in the light of the gained knowledge and critical perspective throughout the investigations. At the end, it is revealed that the smart home has conceptual, technical, and ideological constructive roots within the histories of the modern private sphere, modern architecture and design, and domestic communication and information technologies. It is discussed that these roots constitute the smart home as a domain of social reproduction, which also provides the ground for its conceptualization and promotion as the future domestic sphere.

Η BPL τεχνολογία είναι μια υποσχόμενη τεχνολογία η οποία προσφέρει τηλεπικοινωνιακές υπηρεσίες μέσω των γραμμών ηλεκτρικής ενέργειας και η οποία τα τελευταία χρόνια γνωρίζει ραγδαία ανάπτυξη. Σκοπός της παρούσας διπλωματικής εργασίας είναι η παρουσίαση της BPL τεχνολογίας και η υλοποίηση της με τρεις διαφορετικούς τρόπους σε συγκεκριμένο ηλεκτρικό δίκτυο. Στο πρώτο κεφάλαιο γίνεται μια γενική αναφορά στην BPL τεχνολογία. Παρουσιάζονται τα επίπεδα του δικτύου ηλεκτρικής ενέργειας τα οποία συμμετέχουν σε ένα BPL δίκτυο, τα διάφορα δίκτυα πρόσβασης, οι αρχιτεκτονικές και ο εξοπλισμός τους, καθώς και διάφορα προβλήματα που αντιμετωπίζει η συγκεκριμένη τεχνολογία. Στο δεύτερο κεφάλαιο περιγράφονται τα επίπεδα του BPL δικτύου. Αναλύονται οι διάφορες τεχνικές διαμόρφωσης όπως η OFDM (ορθογώνια μέθοδος διαμόρφωσης με διαίρεση συχνότητας) και η διαμόρφωση απλωμένου φάσματος. Επίσης γίνεται αναφορά στη διαχείρηση σφαλμάτων, στο BPL MAC στρώμα καθώς και στα πρωτόκολλα 802.11. Στο τρίτο κεφάλαιο αναφέρονται οι διάφορες τηλεπικοινωνιακές υπηρεσίες που προσφέρονται από την BPL τεχνολογία καθώς και οι εφαρμογές του έξυπνου δικτύου (Smart Grid) και του έξυπνου σπιτιού (Smart Home) που υλοποιούνται με τη συγκεκριμένη τεχνολογία. Στο τέταρτο κεφάλαιο γίνεται μια σύντομη αναφορά στην τεχνολογία των οπτικών ινών και στο τρόπο με τον οποίο η τεχνολογία αυτή θα χρησιμοποιηθεί ως δίκτυο κορμού το οποίο θα φτάνει μέχρι και τους μετασχηματιστές μέσης/χαμηλής τάσης. Επίσης παρουσιάζεται και η PON-BPL τεχνολογία. Στο πέμπτο και τελευταίο κεφάλαιο πραγματοποιείται η σχεδίαση του BPL δικτύου πρόσβασης με τρεις διαφορετικές αρχιτεκτονικές σε περιοχή της Λακωνίας της οποίας το ηλεκτρικό δίκτυο ψηφιοποιήθηκε με τη χρήση ενός προγράμματος GIS (Γεωγραφικά Συστήματα Πληροφοριών) και συγκεκριμένα του Mapinfo. Στη συνέχεια γίνεται σύγκριση των αρχιτεκτονικών ως προς το κόστος και την ταχύτητα. / BPL technology is a promising technology that offers telecommunications services through power lines, which in the recent years is facing a rapid development. The object of this specific project is the presentation of BPL technology and its realization in three different ways on a specific power grid. In the first chapter there is a general reference to BPL technology. Also it includes a presentation of the levels of the power grid which participate in a specific BPL network. Furthermore there is a presentation of the various access network, their architectures, equipment and various problems faced by this technology. The second chapter contains the description of the levels of BPL network. Also various modulation techniques like OFDM (Orthogonal Frequency Division Multiplexing) and spread-spectrum configuration are analyzed. Finally there is a reference to the management of errors, BPL MAC layer and to the protocols 802.11 as well. In the third chapter there is a reference about the various telecommunication services which are provided via BPL technology as well a reference for the implementations of the Smart Grid and Smart House which are realized with the specific technology. At chapter four a brief reference is conducted about the optical fiber technology and about the manner in which this technology will be used as a backbone which reaches up the medium/low voltage transformers. Also PON-BPL technology is presented. Finally in the fifth and last chapter takes place the design of BPL access network with three different architectures in an area of Laconia whose power grid digitized by using a GIS program (Geographic Information Systems), namely the Mapinfo. Eventually there is a comparison between architectures as concern the cost and the speed of realization needed.

Le Traitement Automatique de la Parole (TAP) s’intéresse de plus en plus et progresse techniquement en matière d’étendue de vocabulaire, de gestion de complexité morphosyntaxique, de style et d’esthétique de la parole humaine. L’Affective Computing tend également à intégrer une dimension « émotionnelle » dans un objectif commun au TAP visant à désambiguïser le langage naturel et augmenter la naturalité de l’interaction personne-machine. Dans le cadre de la robotique sociale, cette interaction est modélisée dans des systèmes d’interaction, de dialogue, qui tendent à engendrer une dimension d’attachement dont les effets doivent être éthiquement et collectivement contrôlés. Or la dynamique du langage humain situé met à mal l’efficacité des systèmes automatiques. L’hypothèse de cette thèse propose dans la dynamique des interactions, il existerait une « glu socio-affective » qui ferait entrer en phases synchroniques deux individus dotés chacun d’un rôle social impliqué dans une situation/contexte d’interaction. Cette thèse s'intéresse à des dynamiques interactionnelles impliquant spécifiquement des processus altruistes, orthogonale à la dimension de dominance. Cette glu permettrait ainsi de véhiculer les événements langagiers entre les interlocuteurs, en modifiant constamment leur relation et leur rôle, qui eux même viennent à modifier cette glu, afin d’assurer la continuité de la communication. La seconde hypothèse propose que la glu socio-affective se construise à partir d’une « prosodie socio-affective pure » que l’on peut retrouver dans certaines formes de micro-expressions vocales. L’effet de ces événements langagiers serait alors graduel en fonction du degré de contrôle d’intentionnalité communicative qui s’observerait successivement par des primitives langagières : 1) des bruits de bouche (non phonétiques, non phonologiques), 2) des sons prélexicaux, 3) des interjections/onomatopées, 4) des imitations à contenu lexical contrôlé. Une méthodologie living-lab est ainsi développée au sein de la plateforme Domus, sur des boucles agiles et itératives co-construites avec les partenaires industriels et sociétaux. Un Magicien d’Oz – EmOz – est utilisé afin de contrôler les primitives vocales comme unique support langagier d’un robot majordome d’un habitat intelligent interagissant avec des personnes âgées en isolement relationnel. Un large corpus, EmOz Elderly Expressions –EEE– est ainsi recueilli. Cet isolement relationnel permet méthodologiquement d’appréhender les dimensions de la glu socio-affective, en introduisant une situation contrastive dégradée de la glu. Les effets des primitives permettraient alors d’observer les comportements de l’humain à travers des indices multimodaux. Les enjeux sociétaux abordés par la gérontechnologie montrent que l’isolement est un facteur de fragilisation où la qualité de la communication délite le maillage relationnel des personnes âgées alors que ces liens sont bénéfiques à sa santé et son bien-être. L’émergence de la robotique d’assistance en est une illustration. Le système automatisé qui découlera des données et des analyses de cette étude permettrait alors d’entraîner les personnes à solliciter pleinement leurs mécanismes de construction relationnelle, afin de redonner l’envie de communiquer avec leur entourage humain. Les analyses du corpus EEE recueilli montrent une évolution de la relation à travers différents indices interactionnels, temporellement organisés. Ces paramètres visent à être intégrés dans une perspective de système de dialogue incrémental – SASI. Les prémisses de ce système sont proposées dans un prototype de reconnaissance de la parole dont la robustesse ne dépendra pas de l’exactitude du contenu langagier reconnu, mais sur la reconnaissance du degré de glu, soit de l’état relationnel entre les locuteurs. Ainsi, les erreurs de reconnaissance tendraient à être compensées par l’intelligence socio-affective adaptative de ce système dont pourrait être doté le robot. / The Natural Language Processing (NLP) has technically improved regarding human speech vocabulary extension, morphosyntax scope, style and aesthetic. Affective Computing also tends to integrate an “emotional” dimension with a common goal shared with NLP which is to disambiguate the natural language and increase the human-machine interaction naturalness. Within social robotics, the interaction is modelled in dialogue systems trying to reach out an attachment dimension which effects need to an ethical and collective control. However, the situated natural language dynamics is undermining the automated system’s efficiency, which is trying to respond with useful and suitable feedbacks. This thesis hypothesis supposes the existence of a “socio-affective glue” in every interaction, set up in between two individuals, each with a social role depending on a communication context. This glue is so the consequence of dynamics generated by a process which mechanisms rely on an altruistic dimension, but independent of dominance dimension as seen in emotions studies. This glue would allow the exchange of the language events between interlocutors, by regularly modifying their relation and their role, which is changing themselves this glue, to ensure the communication continuity. The second hypothesis proposes the glue as built by “socio-affective pure prosody” forms that enable this relational construction. These cues are supposed to be carried by hearable and visible micro-expressions. The interaction events effect would also be gradual following the degree of the communication’s intentionality control. The graduation will be continuous through language primitives as 1) mouth noises (neither phonetics nor phonological sounds), 2) pre-lexicalised sounds, 3) interjections and onomatopoeias, 4) controlled command-based imitations with the same socio-affective prosody supposed to create and modify the glue. Within the Domus platform, we developed an almost living-lab methodology. It functions on agile and iterative loops co-constructed with industrial and societal partners. A wizard of oz approach – EmOz – is used to control the vocal primitives proposed as the only language tools of a Smart Home butler robot interacting with relationally isolated elderly. The relational isolation allows the dimensions the socio-affective glue in a contrastive situation where it is damaged. We could thus observe the primitives’ effects through multimodal language cues. One of the gerontechnology social motivation showed the isolation to be a phenomenon amplifying the frailty so can attest the emergence of assistive robotics. A vicious circle leads by the elderly communicational characteristics convey them to some difficulties to maintain their relational tissue while their bonds are beneficial for their health and well-being. If the proposed primitives could have a real effect on the glue, the automated system will be able to train the persons to regain some unfit mechanisms underlying their relational construction, and so possibly increase their desire to communicate with their human social surroundings. The results from the collected EEE corpus show the relation changes through various interactional cues, temporally organised. These denoted parameters tend to build an incremental dialogue system in perspectives – SASI. The first steps moving towards this system reside on a speech recognition prototype which robustness is not based on the accuracy of the recognised language content but on the possibility to identify the glue degree (i.e. the relational state) between the interlocutors. Thus, the recognition errors avoid the system to be rejected by the user, by tempting to be balanced by this system’s adaptive socio-affective intelligence.

Les habitats intelligents, résultants de la convergence de la domotique, de l'informatique ubiquitaire et de l'intelligence artificielle, assistent leurs habitants dans les situations du quotidien pour améliorer leur qualité de vie.En permettant aux personnes dépendantes et âgées de rester à domicile plus longtemps, ces habitats permettent de fournir une première réponse à des problèmes de société comme la dépendance due au vieillissement de la population.En nous plaçant dans un habitat contrôlé par la voix, l'habitat doit répondre aux requêtes d’un utilisateur concernant un ensemble d’actions pouvant être automatisées (contrôle des lumières, des volets, des dispositifs multimédia, etc.).Pour atteindre cet objectif, le système de contrôle de l'habitat a besoin de prendre en compte le contexte dans lequel un ordre est donné mais également de connaitre les habitudes et préférences de l’utilisateur.Pour cela, le système doit pouvoir agréger les informations issues du réseau de capteurs domotiques hétérogènes et prendre en compte le comportement (variable) de l'utilisateur.La mise au point de systèmes de contrôle intelligent d'un habitat est particulièrement ardue du fait de la grande variabilité concernant aussi bien la topologie des habitats que les habitudes des utilisateurs.Par ailleurs, l'ensemble des informations contextuelles doivent être représentées dans un référentiel commun dans un objectif de raisonnement et de prise de décision.Pour répondre à ces problématiques, nous proposons de développer un système qui d'une part modifie continuellement son modèle de manière à s'adapter à l'utilisateur, et qui d'autre part utilise directement les données issues des capteurs à travers une représentation graphique. L'intérêt et l'originalité de cette méthode sont de ne pas nécessiter d'inférence pour déterminer le contexte.Notre système repose ainsi sur une méthode d'apprentissage par renforcement profond qui couple un réseau de neurones profond du type convolutif permettant l'extraction de données contextuelles, avec un mécanisme d'apprentissage par renforcement pour la prise de décision.Ce mémoire présente alors deux systèmes, un premier reposant uniquement sur l'apprentissage par renforcement et montrant les limites de cette approche sur des environnements réels pouvant comporter plusieurs milliers d'états possibles.L'introduction de l'apprentissage profond a permis la mise au point du second système, ARCADES, dont les bonnes performances montrent la pertinence d'une telle approche, tout en ouvrant de nombreuses voies d'améliorations. / Smart-homes, resulting of the merger of home-automation, ubiquitous computing and artificial intelligence, support inhabitants in their activity of daily living to improve their quality of life.Allowing dependent and aged people to live at home longer, these homes provide a first answer to society problems as the dependency tied to the aging population.In voice controlled home, the home has to answer to user's requests covering a range of automated actions (lights, blinds, multimedia control, etc.).To achieve this, the control system of the home need to be aware of the context in which a request has been done, but also to know user habits and preferences.Thus, the system must be able to aggregate information from a heterogeneous home-automation sensors network and take the (variable) user behavior into account.The development of smart home control systems is hard due to the huge variability regarding the home topology and the user habits.Furthermore, the whole set of contextual information need to be represented in a common space in order to be able to reason about them and make decisions.To address these problems, we propose to develop a system which updates continuously its model to adapt itself to the user and which uses raw data from the sensors through a graphical representation.This new method is particularly interesting because it does not require any prior inference step to extract the context.Thus, our system uses deep reinforcement learning; a convolutional neural network allowing to extract contextual information and reinforcement learning used for decision-making.Then, this memoir presents two systems, a first one only based on reinforcement learning showing limits of this approach against real environment with thousands of possible states.Introduction of deep learning allowed to develop the second one, ARCADES, which gives good performances proving that this approach is relevant and opening many ways to improve it.