Recherches

Jeux de données

• Positions accompagnant Tracking objects within a smart home
• Trackings accompagnant Tracking objects within a smart home

Thèse de doctorat

1. BMP: UN PROTOCOLE DE COMMUNICATION BASÉ SUR LA DISTANCE ENTRE LES OBJETS DE L’INTERNET DES OBJETS (mention: Très Bien)[PDF] [Présentation]

   La quatrième révolution technologique est en marche et stimule des avancées majeures dans les domaines de l’intelligence artificielle et de l’Internet des objets. Cette thèse s’intéresse aux communications entre les objets intelligents. Une revue de la littérature scientifique sur le sujet permet de constater que les protocoles créés pour ces communications s’appuient sur les méthodes de fonctionnement établies à l’époque de la communication entre postes informatiques fixes. Cette thèse propose un nouveau protocole de communication abandonnant ces idées pour plutôt s’appuyer sur la notion de position des objets. Le protocole se nomme BMP, pour Bounded Message Protocol. Les caractéristiques principales du protocole sont que tous les messages sont envoyés en mode diffusion et que la propagation s’arrête lorsqu’une distance du point d’émission est atteinte. Les messages de ce protocole présentent aussi une durée de vie au bout de laquelle chaque message doit être détruit. Ce protocole est conçu pour être léger avec un en-tête minimal à son fonctionnement. Deux implémentations sont réalisées. Une première est en C++ et est utilisée sur des microcontrôleurs de type Arduino. L’expérience réalisée avec cette implémentation permet de valider le bon fonctionnement des mécanismes de contrôle de BMP. Une deuxième implémentation est en Java et est utilisée sur des tablettes et téléphones Android. L’implémentation est utilisée pour un scénario imitant la vie réelle dans un appartement intelligent et confirme que BMP fonctionne dans ce type d’environnement. Parallèlement à BMP, cette thèse présente IPADL (pour Indoor Positioning for Activities of Daily Living), une méthode de positionnement d’objets devant permettre l’usage de BMP à l’intérieur des bâtiments. IPADL utilise des arbres de décision pour convertir un vecteur de puissances de signal issues d’antennes RFID en une position approximative. La méthode est améliorée par l’addition de mesures statistiques sur les puissances.

Articles de journaux révisés par les pairs

1. Bergeron F, Bouchard K, Gaboury S, Giroux S. Tracking objects within a smart home. Expert Systems with Applications, 2018. (IF: 3.928) [PDF] [Positions] [Trackings]

   This paper presents a novel indoor tracking system built with common data mining techniques on radio frequency identification (RFID) tags readings. The system allows tracking of several objects in real-time in a smart home context and is a building block toward the deployment of an expert system to enable aging in place through technology. The indoor localization is modelled as a classification problem, instead of a regression problem as commonly seen in the literature. The paper is divided in two parts. The first one focuses on the ground truth collection that led to the model construction. The second part focuses on the filters that were designed to enable this model to be used in real-time in the smart home as a tracking software. Results from the first part show that most classifiers perform well on the static positioning of RFID tags task, with a random forest of 100 trees performing best at 97% accuracy and 0,974 F-Measure. However, collecting data to train the classifier is a long and tedious process. Results from the second part indicate that the accuracy of the random forest drops significantly when confronted with human interference. With the help of some filters, the tracking accuracy of objects can still be as high as 75%. Those results confirm that using passive RFID tags for an indoor tracking system is viable. Our system is easy to deploy and more flexible than trilateration or fingerprinting systems.

2. Bergeron F, Bouchard K, Gaboury S, Giroux S. RFID Indoor Localization Using Statistical Features. Cybernetics and Systems, 2019 (IF: 1.681) [PDF]

   In this paper, we present a method that uses the signal strength indication of RFID antennas with statistical features to perform relative positioning in a smart home. The goal of the proposed method is to enable the tracking of most objects inside a smart home in real-time, allowing activity recognition based on this tracking. This paper also introduce a new dataset of 4 100 000 RFID data collected in a real full-scale smart home setting. The dataset is available for the community. The method has an accuracy of 95.5% which is similar to previous work but require a fifth of the time to compute.

Articles de conférence révisés par les pairs

1. Bergeron F, Bouchard K, Giroux S, Gaboury S. BMP: A Bounded Message Protocol for the IoT. Proceedings of the 5th EAI International Conference on Smart Objects and Technologies for Social Good, 2019. ACM. [PDF] [Présentation]

   In this paper, we introduce a new communication protocol for machine to machine communication within the Internet of Things (IoT) based on physical distances rather than on addresses. A common challenge with the IoT is to filter the quantity of data that can be collected, especially in centralized solutions. By using distance from the emission point instead of addresses to move messages between objects, this protocol forces the use of decentralized algorithms. It also tackles the problem of the quantity of data by giving each data a time to live, after which messages are deleted by the protocol. The idea is that an object that sends a message should decide who can access it and for how long the data is relevant. We specify the conditions under which such a protocol can be useful and further provide some examples of application that could gain from using this protocol.

2. Bergeron F, Bouchard K, Gaboury S, Giroux S, Bouchard B. Indoor Positioning System for Smart Homes based on Decision Trees and Passive RFID, Pacific Asia Knowledge Discovery and Data Mining Conference, 2016. (17% d'acceptation) [PDF] [Présentation]

   This paper presents a novel Indoor Positioning System (IPS) for objects of daily life equipped with passive RFID tags. The goal is to provide a simple to use, yet accurate, qualitative IPS for housing enhanced with technology (sensors, effectors, etc.). With such a service, the housing, namely called smart home, could enable a wide range of services by being able to better understand the context and the current progression of activities of daily living. The paper shows that classical data mining techniques can be applied to raw data from RFID readers and passive tags. In particular, it explains how we built several datasets using a tagged object in a real smart home infrastructure. Our method was proven very effective as most algorithms result in high accuracy for the majority of the smart home.

3. Bergeron F, Bouchard K, Gaboury S, Giroux S, Bouchard B. Qualitative Tracking of Objects in a Smart Home, PErvasive Technologies Related to Assistive Environments (PETRA) conference, 2016. [PDF] [Présentation]

   This paper presents a novel Indoor Tracking System (ITS) based on passive radio-frequency identification (RFID) technology. The new ITS exploits decision trees built from one dataset per room of a smart home. The datasets are built using a bottle equipped with four class 3 RFID tags and by dividing each room into qualitative zones. The paper discusses how to exploit positioning from decision trees to implement real-time tracking. The long term goal of this ITS is to extract qualitative spatial information to improve recognition of daily living activities' granularity. The results obtained are very encouraging as the average accuracy of the trajectories recognized is over 75%.

Ateliers révisés par les pairs

1. Bergeron F, Bouchard K, Gaboury S, Giroux S, Bouchard B. Simple Objects Tracking System for Smart Homes, IEEE International Conference on Bioinformatics and Biomedicine, 2015. [PDF] [Présentation]

   One of the greatest challenge of research on smart homes is to be able to recognize the ongoing Activity of Daily Living (ADL) in real-time and make prediction on the future action of the resident. To accomplish this task, it is important to have accurate information. In this paper, we present a novel passive RFID Indoor Tracking System (ITS). The goal of this ITS was to create a simple solution from data mining that could be easily deployed in any smart environment without requiring specialized human expertise on RFID. The tracking average 80% of accuracy and show promising results for large scale deployment.

Chapitres de livre collectif

1. Ayimdji Tekemetieu A. , Haidon C. , Bergeron, F., Kenfack Ngankam H., Pigot H., Gouin-Vallerand C., et Giroux S., « Context Modelling in Ambient Assisted Living : Trends and Lessons, » dans Internet of Things - Cases and studies, F. Garca Marquez et B. Lev, éditeurs. Springer, 2021. [PDF]

   The current Internet of Things (IoT) development involves ambient intelligence which ensures that IoT applications provide services that are sensitive, adaptive, autonomous, and personalized to the users’ needs. A key issue of this adaptivity is context modelling and reasoning. Multiple proposals in the literature have tackled this problem according to various techniques and perspectives. This chapter provides a review of context modelling approaches, with a focus on services offered in Ambient Assisted Living (AAL) systems for persons in need of care. We present the characteristics of contextual information, services offered by AAL systems, as well as context and reasoning models that have been used to implement them. A discussion highlights the trends emerging from the scientific literature to select the most appropriate model to implement AAL systems according to the collected data and the services provided.

2. Bouchard K., Bergeron F., Giroux S.: Applying Data Mining in Smart Home, Assistive Technologies in Smart Environments, CRC press, Taylor & Francis, 2015. [PDF]

   This chapter introduces the main data-mining approaches and the works of researchers that applied them in the smart home context. It provides understandable material to exploit data mining for activity recognition in a smart home. The chapter discusses the subject to be used as a basis for further development and research. It discusses the general knowledge related to data mining. The chapter defines what it is and describes the main challenges related to research and application of data mining. The principles are reviewed along with the main limitations for their application in smart home research. The chapter covers the association rules on mining algorithms. It reviews the basic principles and the main works related to smart home research. The chapter investigates clustering for activity recognition. It explains the main algorithms and provides examples of their applications.

Présentation par affiche révisées par les pairs

1. Bergeron F, Bouchard K, Gaboury S, Giroux S. RFID based activities of daily living recognition, 14th IEEE International Conference on Ubiquitous Intelligence and Computing (UIC), 2017. [PDF] [Présentation]

   In this paper, we address the issue of indoor activities of daily living recognition with a novel Activity Recognition System (ARS). This system only uses interactions between objects. Their locations is provided by a tracking system based on passive RFID tags to compute activity probabilities. Classification within the tracking system is done through a random forest that gives over 97% in accuracy. Activities are represented by the use of a custom behaviour tree, which makes it possible to compose interactions to form activities of any complexity. Interactions and activities are defined in a human readable form to allow everyone to expand them with minimal prior knowledge.

2. Bergeron F, Bouchard K, Gaboury S, Giroux S, Bouchard B. Qualitative RFID Tracking for ADL Recognition, AAAI-16 Workshop on artificial intelligence applied to assistive technologies and smart environments, 2016. (17% d'acceptation) [PDF]

   In this paper, we propose a novel approach to perform activities of daily living (ADL) recognition using a tracking system based on the radio-frequency identification (RFID) technology.

3. Bergeron F, Bouchard K, Gaboury S, Giroux S, Bouchard B. Real-Time Constraints for Activities of Daily Living Recognition, PErvasive Technologies Related to Assistive Environments (PETRA) conference, 2016. [PDF] [Affiche]

   In this paper, we present real-time constraints we have to face while building a system to recognize activities of daily living in a smart home using Radio Frequency IDentification (RFID) tag. We also present how we addressed them.

Autres communications

1. Bergeron F, Communication entre objets intelligents, 88e congrès de l'Acfas, mai 2021 [Présentation vidéo]
2. Bergeron F, Communication entre objets intelligents, Midi-conférences JDIs, novembre 2019 [Présentation]
3. Bergeron F, Communication entre objets intelligents, Kaféfak, novembre 2019 [Présentation]
4. Bergeron F, Systèmes de géolocalisation, Midi-conférences JDIs, juillet 2018
5. Bergeron F, Maison, peux-tu m’aider?, Bistro Brain, mai 2018
6. Bergeron F, Maison, peux-tu m’aider?, TechnoDrink, mai 2017
7. Bergeron F, Bouchard K, Gaboury S, Giroux S, Bouchard B. Positionnement intérieur pour la reconnaissance d'activité de la vie quotidienne, Colloque interfacultaire des cycles supérieurs, 2016
8. Bergeron F, Moteurs de recherche, Midi-conférences JDIs, 2016