Les avantages des véhicules autonomes sont formidables, mais le chemin vers une vraie autonomie sera long et semé d’incertitudes. La recherche de ces dernières années s’est basée sur des systèmes multi-capteurs capables de percevoir l’environnement dans lequel le véhicule est conduit. Ces systèmes deviennent plus complexes quand on contrôle le véhicule autonome, différents systèmes de contrôle sont activés dépendant de la décision du système multi-capteurs. Chacun de ces systèmes suit des critères de performance et de stabilité lors de leur conception. Cependant, ils doivent fonctionner ensemble, garantissant une stabilité et étant capable de se charger des changements dynamiques, structuraux et environnementaux. Cette thèse explore la paramétrisation Youla-Kucera (YK) dans des systèmes dynamiques comme les voitures, en insistant sur la stabilité quand la dynamique change, ou que le trafic impose une reconfiguration du contrôleur. Concentrons-nous sur l’obtention de résultats de simulation et expérimentaux en relation avec le "Cooperative Adaptive Cruise Control" (CACC), dans le but, non pas d’utiliser, ici, pour la première fois la paramétrisation YK dans le domaine des systèmes de transport intelligents (STI), mais d’améliorer l’état de l’art en CACC aussi. Des résultats de reconfiguration stable de contrôleurs sont donnés quand la communication avec le véhicule précédent n’est plus disponible, en cas de manœuvre d’entrées/sorties ou lorsqu’ils sont entourés de véhicules aux dynamiques différentes. Ceci démontrant l’adaptabilité, la stabilité et l’implémentation réelle de la paramétrisation YK comme structure générale de contrôle pour les véhicules autonomes. / Benefits of autonomous vehicles are genuinely exciting, but the route to true autonomy in transportation will likely be long and full of uncertainty. Research on the last years is on the development of multi-sensor systems able to perceive the environment in which the vehicle is driving in. These systems increase complexity when controlling an autonomous vehicle, as different control systems are activated depending on the multi-sensor decision system. Each of these systems follows performance and stability criteria for its design, but they all must work together, providing stability guarantees and being able to handle dynamics, structural and environmental changes. This thesis explores the Youla-Kucera (YK) parameterization in dynamics systems such as vehicles, with special emphasis on stability when some dynamics change or the traffic situation demands controller reconfiguration. Focus is in obtaining simulation and experimental results related to Cooperative Adaptive Cruise Control (CACC), with the aim not only of using for the very first time YK parameterization in the Intelligent Transportation Systems (ITS) domain, but improving CACC state-of-the art. Stable controller reconfiguration results are given when non-available communication link with the preceding vehicle, cut-in/out maneuvers or surrounding vehicles with different dynamics, proving adapability, stability and possible real implementation of the YK parameterization as general control framework for autonomous vehicles.

Le véhicule autonome est un véhicule qui se conduira, à terme, sans aucune intervention du conducteur, quelle que soit la situation de conduite. Ce véhicule comprend une nouvelle fonction, nommée fonction AD, pour Autonomous Driving, en charge de la conduite autonome. Cette fonction peut se trouver dans des états différents (Active, Disponible par exemple) selon l'évolution des conditions environnementales. Le changement de ses états est géré par une fonction de Supervision, nommée Supervision AD. Le principal objet de ces travaux consiste à garantir que la fonction AD se trouve constamment dans un état sûr. Ceci revient à s'assurer que la Supervision AD respecte l'ensemble des exigences fonctionnelles et de sûreté qui spécifient son comportement. Ces deux types d'exigences sont émis par deux métiers distincts : l'Architecte Métier Système (AMS) et le pilote Sûreté de Fonctionnement (SdF). Ces deux disciplines d'ingénierie, bien qu'elles contribuent à la conception d'une même fonction, se distinguent en de nombreux points : objectifs, contraintes, planning, outils... Dans notre cas d'étude, ces différences s'illustrent par les exigences considérées : les exigences fonctionnelles sont allouées à la fonction AD globale, tandis que les exigences de sûreté spécifient le comportement de sous-fonctions locales redondantes assurant une continuité de service en cas de défaillance. La mise en cohérence de ces deux perspectives métier au plus tôt dans le cycle de conception et dans un contexte industriel, est la problématique centrale traitée. Les enjeux de SdF soulevés par le véhicule autonome rendent ce problème primordial pour les constructeurs automobiles. Afin de répondre à ces préoccupations, nous avons proposé une démarche outillée et collaborative de conception sûre de la Supervision AD. Cette démarche est intégrée dans les processus normatifs en vigueur (normes ISO 15288 et ISO 26262) ainsi que dans les processus de conception internes chez Renault. Elle est fondée sur la vérification formelle par model checking, la composition parallèle d'automates finis et l'expertise métier. Cette démarche prône l'utilisation d'un même formalisme (l'automate à états finis) par les deux métiers pour mener à bien des activités partageant un objectif de modélisation commun : la vérification d'exigences de comportement en phase amont de conception. Une méthode pour traduire les exigences en propriétés formelles et construire les modèles d'état a été déployée. Il en résulte une consolidation progressive des exigences traitées, initialement rédigées en langage naturel. Les potentielles ambigüités, incohérences et incomplétudes sont exhibées et traitées. / The Autonomous Vehicle is meant to drive itself, without any driver intervention, whatever the driving situation. This vehicle includes a new function, called AD, for Autonomous Driving, function. This function can be in different states (Available, Active for example) according to environmental conditions evolution. This states change is managed by a supervision function, named AD Supervision. The main goal of my works consists in guaranteeing that AD function remains always in a safe state. In other words, the AD Supervision must always respect all the functional and safety requirements that specify its behavior. These two requirements types are produced by two different professions: the System Architect (SA) and the Safety Engineer (SE). These two fields contribute to the design of the same function but distinguish at several aspects: objectives, constraints, planning, tools… In our case study, these differences are illustrated by considered requirements: the functional requirements are allocated to global AD function, while the safety requirements specify the behavior of local redundant sub-functions ensuring a continuous service in case of failure. The consistency of the two perspectives as early as possible in the design phase and in an industrial context, is the central problematic addressed. The safety issues due to Autonomous Vehicle make this topic essential for the automotive manufacturers. To meet these concerns, we proposed a tooled and collaborative approach for safe design of AD Supervision. This approach is integrated in the normative processes (standards ISO 26262 and ISO 15288) as well as in the internal design processes at Renault. It is based on formal verification by model checking, parallel composition of finite sate automata and technical expertise. This approach advocates the utilization of a same formalism (state automata) by the two professions to perform activities sharing a common goal: behavior requirements verification in preliminary design phase. A method to translate requirements into formal properties and to build state models has been deployed. The result is a progressive consolidation of treated requirements, initially expressed in free natural language. The potential ambiguities, inconsistencies and incompleteness are exhibited and treated. Two main contributions are in this way illustrated: highlighting of several formal credible (i.e. validated by expertise) specifications from informal requirements; and precise definition of technical expertise role (milestones, planning). However, this reinforcement – in silos – of the two profession viewpoints does not guarantee that they are mutually consistent. Thus, we proposed a convergence method, relying on expertise and on parallel composition of state automata, for the comparison of local and global views.

The self-driving public bus (SDPB) holds the potential to replace human-operated driving with more eco-friendly means and is therefore a valuable mobility solution for our future. The SDPB is based on the innovative technology of autonomous driving, which can only be guaranteed future market success with broad enough user acceptance. This acceptance is thus an essential factor for the growth of SDPB services. In this context, the travel behavior of young people is particularly interesting, as its development will continually demonstrate future mobility behavior trends. However, little research has been conducted regarding the best methods for motivating young people to accept SDPBs as a viable mode of travel. To address this topic, we first conducted a literature review, identifying factors that potentially influence SDPB acceptance. Subsequently, we developed a comprehensive research model based on the life-oriented approach and the technology acceptance model. This conceptualization was validated by a survey of 268 SDPB riders in real-world traffic. The results reveal several novel factors influencing the acceptance of SDPBs, in particular regarding differences in age. Our research contributes to existing research on both the life-oriented and travel behavior approaches by highlighting age differences and their importance in the field. For instance, our findings demonstrate a vital need to account for age differences when deriving policy implications for future mobility solutions.

Es wird ein umfassendes Verkehrskonzept für das urbane Umfeld beschrieben, welches durch den neuen kollaborativen Fahrzeugtyp mit dem Namen goG [go 'dschi] ermöglicht wird. Fahrzeuge dieses Typs können sich nahtlos im bestehenden Straßenverkehr eingliedern, ohne dass daran Anpassungen vorzunehmen sind. Auf Grund seiner kollaborativen Fähigkeiten können goGs aber auch zu Konvois des ÖPNVs formiert werden und annähernd 80% der im Vergleich für PKWs benötigten Parkplatzfläche einsparen. Des weiteren befähigen sie das goG in voller Fahrt von der Straße an eine über der Straße verlaufenden Umlaufseilbahn zu wechseln. Ein gittergleiches Netzwerk solcher Seilbahnen erlaubt das automatische Durchrouten von goGs 6 Meter über dem Straßenniveau, bei einer konstanten Geschwindigkeit von 50 km/h. An Seilbahnknotenpunkten werden durch Manipulatoren die goGs in voller Fahrt von einer Seilbahn in die vorgesehene nächste Seilbahn umgehängt. Durch hängebrückenartige Aufhängung der Umlaufseilbahnen oberhalb bestehender Straßenzüge werden keine zusätzlichen Flächen benötigt. Der Durchsatz einer Fahrbahn vergrößert sich durch die darüber geführte Umlaufseilbahn um mehr als das Doppelte. Ein autonomes Fahren bei maximal 6km/h zur nächsten goG-Linie des ÖPNV bzw. Seilbahn führt zu einer Vielzahl von Synergieeffekten.

<div>From prehistoric man who needs to be aware of the surrounding situations and hunt for food, to modern industry where machines and robots are programmed to explore the environment and accomplish assignments, situation awareness has always been an essential topic to everyone.</div><div><br></div><div>Advanced Driver-Assistance Systems (ADAS) is one of the modern technologies seeking effective solutions for driving safety. It also utilizes situation awareness model to interpret the driver's state in the environment and provide safe driving advice, with the potential to significantly reduce the traffic accident fatalities.</div><div><br></div><div>To enable situation awareness, an intelligent driving system needs to fulfill the following: (1) perceives the traffic elements in the environment, (2) comprehends the spatial-temporal interactions between a driver and other objects, and (3) projects the states of traffic elements to forecast future actions.</div><div><br></div><div>However, each level of situation awareness encounters its unique challenges in driving scenarios, for example, how to perceive vehicles in low-illuminated conditions? How to represent the complicated interactive relations in complicated driving situations? And how to anticipate the temporal dynamics of traffic elements and identify the where the potential risk comes from? To answer these questions, we explore situation awareness model for Advanced Driver-Assistance Systems at 3 levels: Perception, Comprehension and Projection. We discuss how to realize situation awareness based on three different computer vision tasks. We demonstrate that our proposed system is able to forecast the driver's operational intentions and identify risk objects to avoid hazards.</div>

Semantic segmentation is the task of predicting predefined class labels for each pixel in a given image. It is essential in autonomous driving, but also challenging because training accurate models requires large and diverse datasets, which are difficult to collect due to the high cost of annotating images at pixel-level. This raises interest in using synthetic images from simulators, which can be labelled automatically. However, models trained directly on synthetic data perform poorly in real-world scenarios due to the distributional misalignment between synthetic and real images (domain shift). This thesis explores the effectiveness of several techniques for alleviating this issue, employing Synscapes and Cityscapes as the synthetic and real datasets, respectively. Some of the tested methods exploit a few additional labelled real images (few-shot supervised domain adaptation), some have access to plentiful real images but not their associated labels (unsupervised domain adaptation), and others do not take advantage of any image or annotation from the real domain (domain generalisation). After extensive experiments and a thorough comparative study, this work shows the severity of the domain shift problem by revealing that a semantic segmentation model trained directly on the synthetic dataset scores a poor mean Intersection over Union (mIoU) of 33:5% when tested on the real dataset. This thesis also demonstrates that such performance can be boosted by 25:7% without accessing any annotations from the real domain and 17:3% without leveraging any information from the real domain. Nevertheless, these gains are still inferior to the 31:0% relative improvement achieved with as little as 25 supplementary labelled real images, which suggests that there is still room for improvement in the fields of unsupervised domain adaptation and domain generalisation. Future work efforts should focus on developing better algorithms and creating synthetic datasets with a greater diversity of shapes and textures in order to reduce the domain shift. / Semantisk segmentering är uppgiften att förutsäga fördefinierade klassetiketter för varje pixel i en given bild. Det är viktigt för autonom körning, men också utmanande eftersom utveckling av noggranna modeller kräver stora och varierade datamängder, som är svåra att samla in på grund av de höga kostnaderna för att märka bilder på pixelnivå. Detta väcker intresset att använda syntetiska bilder från simulatorer, som kan märkas automatiskt. Problemet är emellertid att modeller som tränats direkt på syntetiska data presterar dåligt i verkliga scenarier på grund av fördelningsfel mellan syntetiska och verkliga bilder (domänskift). Denna avhandling undersöker effektiviteten hos flera tekniker för att lindra detta problem, med Synscapes och Cityscapes som syntetiska respektive verkliga datamängder. Några av de testade metoderna utnyttjar några ytterligare märkta riktiga bilder (few-shot övervakad domänanpassning), vissa har tillgång till många riktiga bilder men inte deras associerade etiketter (oövervakad domänanpassning), och andra drar inte nytta av någon bild eller annotering från den verkliga domänen (domängeneralisering). Efter omfattande experiment och en grundlig jämförande studie visar detta arbete svårighetsgraden av domänskiftproblemet genom att avslöja att en semantisk segmenteringsmodell som upplärts direkt på den syntetiska datauppsättningen ger en dålig mean Intersection over Union (mIoU) på 33; 5% när den testas på den verkliga datamängden. Denna avhandling visar också att sådan prestanda kan ökas med 25; 7% utan att komma åt några annoteringar från den verkliga domänen och 17; 3% utan att utnyttja någon information från den verkliga domänen. Ändå är dessa vinster fortfarande sämre än den 31; 0% relativa förbättringen som uppnåtts med så lite som 25 kompletterande annoterade riktiga bilder, vilket tyder på att det fortfarande finns utrymme för förbättringar inom områdena oövervakad domänanpassning och domängeneralisering. Framtida arbetsinsatser bör fokusera på att utveckla bättre algoritmer och på att skapa syntetiska datamängder med en större mångfald av former och texturer för att minska domänskiftet.

The objective of this project is to develop a driving test using a Digital Human Modeling tool (DHM), specifically IPS IMMA, which will allow the evaluation of the ergonomics of the interior of vehicles as currently demanded by the automotive companies. Thus, improving both the design and the design process. This will involve a study of the driving and the tasks carried out by a real person to end up programming them in the DHM software. Based on this study an interface is suggested that guides engineers or ergonomists to design their own driving tests and enable them to evaluate their own designs without a high specialization in DHM tools and software. Taking into account the already present autonomous cars and their future development, the conceptual design of a two positions steering wheel (autonomous/manual driving) will be introduced as an example to be added in the driving test. This example is intended to show how DHM tools can be used to evaluate different designs solutions in early stages of the product development process. This project will be a contribution to one of the sections of the ADOPTIVE project carried out at the University of Skövde and in collaboration with Swedish automotive companies.

Le déploiement des futurs véhicules autonomes promet d'avoir un impact socio-économique majeur, en raison de leur promesse d'être à la fois plus sûrs et plus efficaces que ceux conduits par des humains. Afin de satisfaire à ces attentes, la capacité des véhicules autonomes à planifier des trajectoires sûres et à manœuvrer efficacement dans le trafic sera capitale. Cependant, le problème de planification de trajectoire au milieu d'obstacles statiques ou mobiles a une combinatoire forte qui est encore aujourd'hui problématique pour les meilleurs algorithmes.Cette thèse explore une nouvelle approche de la planification de mouvement, basée sur l'utilisation de la notion de décision de conduite comme guide pour structurer le problème de planification en vue de faciliter sa résolution. Cette approche peut trouver des applications pour la conduite coopérative, par exemple pour coordonner plusieurs véhicules dans une intersection non signalisée, ainsi que pour la conduite autonome où chaque véhicule planifie sa trajectoire. Dans le cas de la conduite coopérative, les décisions correspondent au choix d'un ordonnancement des véhicules qui peut être avantageusement encodé comme un graphe. Cette thèse propose une représentation similaire pour la conduite autonome, où les décisions telles que dépasser ou non un véhicule sont nettement plus complexes. Une fois la décision prise, il devient aisé de déterminer la meilleure trajectoire y correspondant, en conduite coopérative comme autonome. Cette approche basée sur la prise de décision peut permettre d'améliorer la robustesse et l'efficacité de la planification de trajectoire, et ouvre d'intéressantes perspectives en permettant de combiner des approches mathématiques classiques avec des techniques plus modernes d'apprentissage automatisé. / The deployment of future self-driving vehicles is expected to have a major socioeconomic impact due to their promise to be both safer and more traffic-efficient than human-driven vehicles. In order to live up to these expectations, the ability of autonomous vehicles to plan safe trajectories and maneuver efficiently around obstacles will be paramount. However, motion planning among static or moving objects such as other vehicles is known to be a highly combinatorial problem, that remains challenging even for state-of-the-art algorithms. Indeed, the presence of obstacles creates exponentially many discrete maneuver choices, which are difficult even to characterize in the context of autonomous driving. This thesis explores a new approach to motion planning, based on using this notion of driving decisions as a guide to give structure to the planning problem, ultimately allowing easier resolution. This decision-based motion planning approach can find applications in cooperative driving, for instance to coordinate multiple vehicles through an unsignalized intersection, as well as in autonomous driving where a single vehicle plans its own trajectory. In the case of cooperative driving, decisions are known to correspond to the choice of a relative ordering for conflicting vehicles, which can be conveniently encoded as a graph. This thesis introduces a similar graph representation in the case of autonomous driving, where possible decisions -- such as overtaking the vehicle at a specific time -- are much more complex. Once a decision is made, planning the best possible trajectory corresponding to this decision is a much simpler problem, both in cooperative and autonomous driving. This decision-aware approach may lead to more robust and efficient motion planning, and opens exciting perspectives for combining classical mathematic programming algorithms with more modern machine learning techniques.

No description available.

The presence of electrical and electronic circuits in commercial trucks has increased at a very fast rate during recent decades. With advancements in embedded systems and the introduction of electric controls in the automotive industry, the design of complex electric systems for the vehicles has become one of the major design challenges. In the commercial truck industry, the development cycles are almost a decade long. Therefore, it is a big challenge to introduce a new architecture to accommodate the modern automotive technologies in the upcoming generation of trucks. Currently, the commercial truck industry relies highly on a federated electrical/electronic (E/E) architecture. In this architecture, Electronic Control Units (ECU) are responsible for computation and Input/Output operations. These ECUs are clustered into different domains based on their respective functions. However, these domains are not isolated from each other. These modules communicate with each other using a vehicular network, which is typically a controller area network in the current trucks. In the automotive industry, automation is increasing at a fast pace. As the level of automation increases, the need for high computation also increases, which increases the overall costs. This study aims to address this problem by introducing an integrated E/E architecture where all the computational power is concentrated at one place (or perhaps two or three places to allow for redundancy). This study proposes to introduce a lowcost replacement for the current ECUs with more limited computational power but with generic input/output interfaces. This thesis provides the reader with some background of the current E/E architecture of commercial trucks and introduces the reader to ECUs. Additionally, the relevant network architectures and protocols are explained. A potential solution, based upon the centralized computation based E/E architecture and its implementation are discussed followed by a detailed analysis of the replacements for ECUs. The result of this analysis, if adopted, should result in a reduction of manufacturing and design costs, as well as make the production and maintenance process easier. Moreover, this should also have environmental benefits by reducing fuel consumption. / Förekomsten av elektronik och elektriska kretsar I kommersiella lastbilar has ökat i en väldigt snabb takt under de senaste decennierna. Med framsteg inom inbyggda system och introduktionen av elektroniska styrsystem i fordonsindustrin så har komplexa elektroniska system blivit en av de största designutmaningarna. I den kommersiella lastbilsindustrin där utvecklingscyklerna är nästan ett decennium, är det en stor utmaning att introducera ny arkitektur som tillgodoser all den nya teknologin som införlivas i fordonet. För närvarande så förlitar sig den kommersiella lastbilsindustrin mycket på en federated elektrisk/elektronisk (E/E) arkitektur. I denna arkitektur är elektroniska styrenheter (ECU) ansvariga för beräkningar och I/O (Input/Output) operationer. Dessa ECU:er är samlade i olika domäner baserade på dess funktioner. Domänerna är dock inte isolerade från varandra. De här modulerna kommunicerar därför med varandra med hjälp av ett fordonsnätverk, typiskt en CAN (Controller Area Network) i nuvarande lastbilar. I fordonsindustrin ökar automatiseringen i en snabb fart. I takt med att automatiseringen ökar så ökar även behovet av snabba och energiintensiva beräkningar, vilket i sin tur ökar den totala kostnaden. Denna studie har som mål att adressera det här problemet genom att introducera en integrated E/E arkitektur där all beräkningskraft är koncentrerad till en plats (eller två eller tre platser för att tillåta överskott). Den här studien föreslår att introducera en ersättning av nuvarande ECU:er till en låg kostnad, med lägre beräkningskraft och generiska I/O gränssnitt. Studien föreslår också ersättningar av nuvarande fordonsnätverk. Den här uppsatsen förser läsaren med viss bakgrund till den nuvarande E/E arkitekturen för kommersiella lastbilar och introducerar läsaren till ECU:er. Dessutom förklaras de relevanta nätverksarkitekturerna och protokollen. En potentiell lösning som baseras på den integrated E/E arkitekturen och dess implementering diskuteras med fokus på en detaljerad analys av ersättningarna till ECU:er. Resultatet av den här analysen skulle, om den adopteras, medföra minskning av tillverknings- och designkostnader samt leda till en förenkling av produktion och underhåll. Utöver det så bör det även ha miljöfördelar genom minskad bränsleförbrukning.