L'étude des caractéristiques faciales a suscité l'intérêt croissant de la communauté scientifique et des industriels. En effet, ces caractéristiques véhiculent des informations non verbales qui jouent un rôle clé dans la communication entre les hommes. De plus, elles sont très utiles pour permettre une interaction entre l'homme et la machine. De ce fait, l'étude automatique des caractéristiques faciales constitue une tâche primordiale pour diverses applications telles que les interfaces homme-machine, la science du comportement, la pratique clinique et la surveillance de l'état du conducteur. Dans cette thèse, nous nous intéressons à la surveillance de l'état du conducteur à travers l'analyse de ses caractéristiques faciales. Cette problématique sollicite un intérêt universel causé par le nombre croissant des accidents routiers, dont une grande partie est provoquée par une dégradation de la vigilance du conducteur, connue sous le nom de l'hypovigilance. En effet, nous pouvons distinguer trois états d'hypovigilance. Le premier, et le plus critique, est la somnolence qui se manifeste par une incapacité à se maintenir éveillé et se caractérise par les périodes de micro-sommeil correspondant à des endormissements de 2 à 6 secondes. Le second est la fatigue qui se définit par la difficulté croissante à maintenir une tâche à terme et se caractérise par une augmentation du nombre de bâillements. Le troisième est l'inattention qui se produit lorsque l'attention est détournée de l'activité de conduite et se caractérise par le maintien de la pose de la tête en une direction autre que frontale. L'objectif de cette thèse est de concevoir des approches permettant de détecter l'hypovigilance chez le conducteur en analysant ses caractéristiques faciales. En premier lieu, nous avons proposé une approche dédiée à la détection de la somnolence à partir de l'identification des périodes de micro-sommeil à travers l'analyse des yeux. En second lieu, nous avons introduit une approche permettant de relever la fatigue à partir de l'analyse de la bouche afin de détecter les bâillements. Du fait qu'il n'existe aucune base de données publique dédiée à la détection de l'hypovigilance, nous avons acquis et annoté notre propre base de données représentant différents sujets simulant des états d'hypovigilance sous des conditions d'éclairage réelles afin d'évaluer les performances de ces deux approches. En troisième lieu, nous avons développé deux nouveaux estimateurs de la pose de la tête pour permettre à la fois de détecter l'inattention du conducteur et de déterminer son état, même quand ses caractéristiques faciales (yeux et bouche) ne peuvent être analysées suite à des positions non-frontales de la tête. Nous avons évalué ces deux estimateurs sur la base de données publique Pointing'04. Ensuite, nous avons acquis et annoté une base de données représentant la variation de la pose de la tête du conducteur pour valider nos estimateurs sous un environnement de conduite. / Studying facial features has attracted increasing attention in both academic and industrial communities. Indeed, these features convey nonverbal information that plays a key role in humancommunication. Moreover, they are very useful to allow human-machine interactions. Therefore, the automatic study of facial features is an important task for various applications includingrobotics, human-machine interfaces, behavioral science, clinical practice and monitoring driver state. In this thesis, we focus our attention on monitoring driver state through its facial features analysis. This problematic solicits a universal interest caused by the increasing number of road accidents, principally induced by deterioration in the driver vigilance level, known as hypovigilance. Indeed, we can distinguish three hypovigilance states. The first and most critical one is drowsiness, which is manifested by an inability to keep awake and it is characterized by microsleep intervals of 2-6 seconds. The second one is fatigue, which is defined by the increasing difficulty of maintaining a task and it is characterized by an important number of yawns. The third and last one is the inattention that occurs when the attention is diverted from the driving activity and it is characterized by maintaining the head pose in a non-frontal direction.The aim of this thesis is to propose facial features based approaches allowing to identify driver hypovigilance. The first approach was proposed to detect drowsiness by identifying microsleepintervals through eye state analysis. The second one was developed to identify fatigue by detecting yawning through mouth analysis. Since no public hypovigilance database is available,we have acquired and annotated our own database representing different subjects simulating hypovigilance under real lighting conditions to evaluate the performance of these two approaches. Next, we have developed two driver head pose estimation approaches to detect its inattention and also to determine its vigilance level even if the facial features (eyes and mouth) cannot be analyzed because of non-frontal head positions. We evaluated these two estimators on the public database Pointing'04. Then, we have acquired and annotated a driver head pose database to evaluate our estimators in real driving conditions.

It is estimated that as many as 2.7 million crashes occur each year at intersections or are intersection related; resulting in over 8500 fatalities each year. These statistics have prompted government and corporate sponsored research into collision countermeasure systems that can enhance safety at intersections. Researchers are investigating technologies to provide an infrastructure-based or infrastructure-cooperative Intersection Decision Support (IDS) systems. Such systems would use pre-specified algorithms to identify drivers that have a high likelihood of violating the traffic signal and thus increase the risk of a collision. The system would subsequently warn the violating driver to stop though an in-vehicle or infrastructure-mounted interface. An IDS algorithm must be designed to provide adequate time for the driver to perceive, react, and stop the vehicle, while simultaneously avoiding a high false alarm rate. Prior to developing these algorithms, scientists must understand how drivers respond to traffic signals. Little research has focused on the influence of driver state on red-light running behavior or methods for distinguishing red light violators from non-violators. The objective of the present study was to define trends associated with intersection crossings under different driver states and to explore the point detection method of predicting red light running upstream of the intersection. This was accomplished through a test-track mixed-factor experiment with 28 participants. Each participant experienced a baseline (complete a full stop at the red light), distracted (misses signal phase change due to inattention), and willful (driver knowingly makes a late crossing in an attempt to 'beat the light') driver state conditions. To provide the opportunity for red-light running behavior from participants, the amber change interval began at five different distances from the intersection. These distances were located near and within the dilemma zone, a region in which drivers have a difficult time deciding whether to go or to stop. Data collected from in-vehicle sensors was statistically analyzed to determine significant effects between driver states, and to investigate point detection algorithms. / Master of Science

Driving a heavy-duty distribution truck in an urban environment is an intricate task that requires atremendous effort from the driver. Even a minor error such as briefly looking away from the road for asecond could have fatal consequences. The automotive industry is heavily investing into making roads saferand innovations such as the seat belt has been able to save millions of lives. With the introduction of new lawsand regulation there is an emerging adaptation of driver monitoring systems in vehicles which aims to savemore lives and improve the driving experience. This master thesis presents the exploration and design of a holistic concept for how driver state monitoringcould be implemented in a manual heavy-duty distribution truck in the year of 2035 at Scania. The mainpurpose is to add value to drivers by finding solutions that bridge driver needs and monitoring technology.Driver state monitoring is a field that introduces abilities for the vehicle to connect and understand the driverin order to provide more efficient assists for a safer and more pleasurable driving experience. An iterative usercentered design process was initialized with user interviews, observations and literature reviews that resulted inan in-depth view of the driver’s needs and a summary of the latest research and innovations. A creative phasegenerated multiple ideas that where evaluated using qualitative feedback from drivers and experts. A final setof features was selected and by using 3D modelling and video editing, an eight minute long animated film wascreated that showcases the possibilities of driver state monitoring in a heavy-duty distribution truck in 2035. The final concept is a unified vision of 22 features in a heavy-duty electric distribution truck, utilizing hybriddriver state monitoring techniques in order to provide a safer and more pleasurable driving experience. It isinspired and created around the latest research and innovations in the automotive industry and aims to inspireScania in their future work to implement and innovate new solutions based on driver state monitoring. This isa thesis work by two Master of Science student within Industrial Design Engineering from Luleå University ofTechnology at Scania in the spring of 2020. / Att köra en distributionslastbil i stadsmiljö är en komplicerad uppgift som kräver en stor insats från föraren. Till och med ett litet misstag som att titta bort från vägen under en kort period kan få katastrofala följder.Fordonsindustrin investerar mycket inom säkerhet och innovationer som säkerhetsbältet har väldigtframgångsrikt bidragit till att rädda miljontals liv på vägarna. I och med nya lagar implementeras system ifordon som bevakar föraren, med målet att rädda fler liv och skapa en bättre körupplevelse. Det här examensarbetet syftar till att designa ett helhetskoncept som visar hur bevakning av föraren kanimplementeras i en manuell distributionslastbil på Scania år 2035. Det huvudsakliga målet är att skapa värdeför förarna genom att hitta lösningar som bryggar gapet mellan förarnas behov och tekniken som möjliggörbevakning av föraren. Genom att introducera denna typ av förarbevakningsteknik ges möjligheten förfordonet att bättre förstå sig på föraren, vilket i sin tur möjliggör ökad säkerhet och bättre körupplevelser. Eniterativ och användarcentrerad designprocess användes och inleddes med intervjuer, observationer och enlitteraturstudie, vilket resulterade i en djup förståelse för förarnas behov samt en summering av den senasteforskningen och innovationerna. En kreativ fas genererade en mängd idéer som sedan utvärderas med hjälp avkvalitativ feedback från både förare och experter. Ett slutligt urval av funktioner togs fram och visualiseradesgenom att skapa en animerad film. Den färdiga åtta minuter långa filmen visar på möjligheterna med tekniksom bevakar föraren i en distributionslastbil år 2035. Det slutliga konceptet förenar 22 funktioner i en elektrisk distributionslastbil. Ett flertal tekniker utnyttjas föratt bevaka föraren och tillsammans bidrar de till en säkrare och mer njutbar körupplevelse. Konceptet ärinspirerat av- och bygger på den senaste forskningen och de senaste innovationerna inom fordonsindustrin,med målet att inspirera Scania i framtida projekt där förarbevakningssystem utvecklas och implementeras.Detta arbete är ett examensarbete gjort av två civilingenjörsstudenter inom tekniks design från Luleå tekniskauniversitet på Scania under våren 2020.

Academic studies of sustainable fashion, and the discourses of actors in business and policy, under-define fashion as a system by treating the social and ecological aspects of fashion separately. This reduces the potential for academic findings to provide knowledge useful for transformation of the fashion system and obstructs desired outcomes from policy and business responses to fashion’s negative social and environmental impacts. Understanding how fashion works as a system presents a challenge to transdisciplinary efforts for transformation towards sustainability. In this Licentiate, I explore ways to look at fashion using a feminist critical realist social-ecological system approach. I develop a theoretical framework to understand the fashion system, and particularly to understand what is keeping it unsustainable. I view fashion as a ‘nested’ social-ecological system with inseparable social and biophysical parts.  I use a feminist lens characterized by diversity; this draws attention to gaps, what is known, missing and absent. To show that social aspects and material aspects are intertwined and cannot be studied independently of each other, I use critical realism as a metatheory. I bring its idea of a stratified reality and the model of the four-planar social being to the social-ecological system approach that forms the core of my work. I combine Ostrom’s frequently used general framework for analysing social-ecological systems with a policy-oriented framework for sustainable development. Drawing from these two frameworks I develop a five principles for a strategy framework for sustainable fashion. In summary, applying the strategy framework within the theoretical framework enables thinking more deeply about the structure and implications of knowledge contributions when taking a social-ecological perspective on actions for sustainability.  The two papers in this licentiate thesis examine the effects of ontological standpoints that allow environmental impacts of textile fibres to be analysed in isolation from the cultural and social aspects of fashion.    Paper 1, ‘Making Resilient Decisions for Sustainable Circularity of Fashion’, is recently published in the journal Circular Economy and Sustainability (Palm et al. 2021). It aimed to show how current circularity responses to global sustainability challenges have so far fallen short. The current path of the expanding fashion industry is fraught with accelerated material throughputs and increased disposal and waste, contributing to human-driven environmental changes at planetary scale. In addition the fashion industry has issues of poor working conditions, modern-day slavery, and justice. By representing a Driver – State – Response framework as an adaptive cycle of a social-ecological system, it makes it clear that reducing planetary pressure from the global fashion and textiles industry requires greater recognition of the system’s social drivers. This paper was a step towards the iterative development of my sustainable fashion framework.   Paper 2, ‘Reviewing and defining the concept of Sustainable Fashion: a critical social-ecological approach’, is included as an early-stage draft manuscript. It aims to provide a starting point for discussions towards a coherent science-business-policy definition of the concept of sustainable fashion itself. Using the five theoretically grounded principles of my strategy framework, I examine the manifold definitions related to sustainable fashion such as eco fashion, green fashion, ethical fashion, slow fashion, organic fashion and cradle-to-cradle-fashion. Critical realism’s idea of absence structures this paper.  This thesis contributes to knowledge of what a nested inseparable social-ecological system fashion is, enriching ontological descriptions for resilience research more generally.  Also, it provides concrete guidance for transdisciplinary efforts with business and policy working to decrease fashion’s negative impacts on humans and the planet, by showing that fruitful responses pay attention to social activities beyond the industry value chain, not just material flows within. Finally,  I hope my research serves as a contribution to propaedeutics of the field of sustainable fashion, i.e. giving an introductory understanding of the reality and the possibilities of fashion for people and planet. / <p>The work presented here was part of a consortium research project between Stockholm Resilience Centre and the Ellen MacArthur Foundation, initiated and funded by H&amp;M Group, the Swedish public limited company. Celinda Palm, Sarah Cornell and Tiina Häyhä’s employment was part-funded through this project.</p>