<p>Emerging technologies such as real-time travel information systems and automated vehicles (AVs) have profound impacts on driver decision-making behavior. While they generally have positive impacts by enabling drivers to make more informed decisions or by reducing their driving effort, there are several concerns related to inadequate consideration of cognitive and psychological aspects in their design. In this context, this dissertation analyzes different aspects of driver cognition and psychology that arise from drivers’ interactions with these technologies using physiological data collected in two sets of driving simulator experiments.</p> <p>This research analyzes the latent cognitive and psychological effects of real-time travel information using electroencephalogram (EEG) data measured in the first set of driving simulator experiments. Using insights from the previous analysis, a hybrid route choice modeling framework is proposed that incorporates the impacts of the latent information-induced cognitive and psychological effects along with other explanatory variables that can be measured directly (i.e., route characteristics, information characteristics, driver attributes, and situational factors) on drivers’ route choice decisions. EEG data is analyzed to extract two latent cognitive variables that capture the driver’s cognitive effort during and immediately after the information provision, and cognitive inattention before implementing the route choice decision. </p> <p>Several safety concerns emerge for the transition of control from the automated driving system to a human driver after the vehicle issues a takeover warning under conditional vehicle automation (SAE Level 3). In this context, this study investigates the impacts of driver’s pre-warning cognitive state on takeover performance (i.e., driving performance while resuming manual control) using EEG data measured in the second set of driving simulator experiments. However, there is no comprehensive metric available in the literature that could be used to benchmark the role of driver’s pre-warning cognitive state on takeover performance, as most existing studies ignore the interdependencies between the associated driving performance indicators by analyzing them independently. This study proposes a novel comprehensive takeover performance metric, Takeover Performance Index (TOPI), that combines multiple driving performance indicators representing different aspects of takeover performance. </p> <p>Acknowledging the practical limitations of EEG data to have real-world applications, this dissertation evaluates the driver’s situational awareness (SA) and mental stress using eye-tracking and heart rate measures, respectively, that can be obtained from in-vehicle driver monitoring systems in real-time. The differences in SA and mental stress over time, their correlations, and their impacts on the TOPI are analyzed to evaluate the efficacy of using eye-tracking and heart rate measures for estimating the overall takeover performance in conditionally AVs.</p> The study findings can assist information service providers and auto manufacturers to incorporate driver cognition and psychology in designing safer real-time information and their delivery systems. They can also aid traffic operators to incorporate cognitive aspects while devising strategies for designing and disseminating real-time travel information to influence drivers’ route choices. Further, the study findings provide valuable insights to design operating and licensing strategies, and regulations for conditionally automated vehicles. They can also assist auto manufacturers in designing integrated in-vehicle driver monitoring and warning systems that enhance road safety and user experience.

<div>In this thesis, we explore different methods to enhance the safety and robustness for autonomous systems. We achieve this goal using concepts and tools from reachability analysis and control barrier functions. We first take on a multi-player reach-avoid game that involves two teams of players with competing objectives, namely the attackers and the defenders. We analyze the problem and solve the game from the attackers' perspectives via a moving horizon approach. The resulting solution provides a safety guarantee that allows attackers to reach their goals while avoiding all defenders. </div><div><br></div><div>Next, we approach the problem of target re-association after long-term occlusion using concepts from reachability as well as Bayesian inference. Here, we set out to find the probability identity matrix that associates the identities of targets before and after an occlusion. The solution of this problem can be used in conjunction with existing state-of-the-art trackers to enhance their robustness.</div><div><br></div><div>Finally, we turn our attention to a different method for providing safety guarantees, namely control barrier functions. Since the existence of a control barrier function implies the safety of a control system, we propose a framework to learn such function from a given user-specified safety requirement. The learned CBF can be applied on top of an existing nominal controller to provide safety guarantees for systems.</div>

No description available.

This thesis project aims to target the increasing number of people who live, work and transit through the densely populated metropolises that, fused together, create megaregions. These individuals, an already big and ever increasing number of people, are the so-called super-commuters, members of the workforce whose commutes surpass the figures of 90 minutes or, alternatively, 145 km in a single-way.As it happens with others who live in the same geographical region, they experience the need or wish of working in the dense urban centres that offer plenty of job opportunities, but little housing opportunities. This lack of housing paired with its very high and ever increasing prices forces these people to disperse and to search for living places farther away from their workplaces, thus trading convenience and free time for long commutes. Super-commuting is indeed showing a growing trend, not only in the number of super-commuters themselves but also in the duration of commutes.Commutes are getting lengthier for a number of reasons, such as increasingly comfortable vehicles, technological advancements that help render commuting time either productive or entertaining.However, in a post-pandemic society, many companies are also offering their employees the chance to adopt hybrid work modes with more days spent working from home - which makes workers consider living farther away from their workplaces as they need to commute less (number of times). It is an undeniable fact that the longer the commute, the less free time one worker has, either for resting or doing something productive. It is also true that in an increasingly fast-paced technological world, people have also increasing difficulties in separating their professional and personal lives. Therefore, the approach of this thesis project goes through offering people the chance to make the most out of their commutes, so as to free more of their time when not commuting or working - time to spend with their loved ones or to be used to do whatever they would like. To achieve that, this project contemplates the use of autonomous technologies expected to become more widespread within the automotive industry; as by rendering vehicles autonomous would free people from driving and let them allocate their time to other tasks. This thesis project offers a holistic proposal of a premium commuting service targeted to super-commuters living and working within the Northern California megaregion. This service would connect peripheral communities directly to the Bay Area, where most companies are located, through a door-to-hub service.It focuses on how users of this service might experience their commutes by presenting case studies of three different types of professionals with diverse needs, and exploring how they would use it as a means of making their commutes as convenient and efficient as possible. The process herein exposed goes through the various stages of design development, from research to ideation and leading to a final proposal, consisting of a service, mobile booking app and exterior + interior design of a vehicle.

Utvecklingen inom fordonsindustrin går mot autonoma fordon, och som effekt av det behöver både säkerhet och samhällets acceptans öka. För att öka säkerheten i autonoma fordon behöver föraren vara situationsmedveten, vilket kan uppnås genom en röstassistent i fordonet. Om säkerheten skulle öka skulle sannolikt även människors tillit och därav samhällets acceptans öka. Studier pekar på att olika egenskaper så som rösttyp hos röstassistenter kan ha en effekt på förares situationsmedvetenhet och tillit. Denna studie ämnar således att undersöka huruvida en feminin eller maskulin röst hos röstassistenter påverkar förares situationsmedvetenhet och upplevda tillit till autonoma fordon. För att ta reda på detta genomfördes en experimentell studie där deltagare befann sig i en bilsimulator av ett autonomt fordon som simulerade körscenarier. Deltagarna var uppdelade i två grupper där varje grupp fick uppleva en typ av röstassistent, antingen med feminin eller maskulin röst. Deltagarna fick uppleva körscenarier där varningssituationer i olika allvarlighetsgrader uppstod, där de fick fatta beslut kring hur de hade hanterat varningssituationen. I samband med körscenarierna utfördes tre olika datainsamlingar bestående av intervju och enkäter. Resultatet av studien påvisar ingen signifikant eller utmärkande skillnad vad gäller upplevd tillit mellan grupper. Resultatet indikerar inte heller någon utmärkande skillnad mellan gruppernas situationsmedvetenhet. Det finns marginell skillnad mellan grupperna, där gruppen som upplevde feminin röst hade både högre upplevd tillit och situationsmedvetenhet. Slutresultatet i denna studie visar att området behöver utforskas vidare för att upptäcka och förstå eventuella effekter av rösttyper hos röstassistenter. / The development in the automotive industry is moving towards autonomous vehicles, and as a consequence the need to increase both safety and societal acceptance is necessary. To increase safety in autonomous vehicles, the driver needs to be situationally aware, which can be achieved through a voice assistant in the vehicle. If safety were to increase, people's trust and therefore society's acceptance would most likely also increase. Studies indicate that different characteristics of voice assistants, such as the type of voice, can have an effect on drivers' situational awareness and trust. Thus, this study aims to investigate whether a feminine or masculine voice in voice assistants affects drivers' situational awareness and perceived trust in autonomous vehicles. An experimental study was conducted in order to investigate this, where participants were situated in a simulator of an autonomous vehicle that simulated driving scenarios. The participants were divided into two groups where each group experienced one of two types of voice assistants, either with a feminine or a masculine voice. The participants had to experience driving scenarios where warning situations of varying degrees of severity arose, where they had to make decisions about how they would have handled the situation. In connection with the driving scenarios, three types of data collections consisting of interviews and questionnaires were carried out. The results of the study show no significant or distinctive difference in terms of perceived trust between groups. The results also don’t indicate any distinctive difference between the groups' situational awareness. There is marginal difference between the groups, with the group experiencing feminine voice having both higher perceived trust and situational awareness. The final result of this study shows that this research area needs to be explored further to discover and understand the possible effects of voice types in voice assistants.

In the nearest future, the human driver is viewed as a reliable backup even for the fully automated road motor vehicles (cars). Indeed, the driver is assumed to swiftly take the control of the car in cases of suddenly occurring (i) challenging environmental conditions, (ii) complex unforeseen driving situations, or (iii) degradation of performance of the car. However, due to the cognitive overload in such a sudden, stressful takeover of the control, the driver would often experience the startle effect, which usually results in an unconscious, instinctive, yet incorrect response. An extreme case of startle is freezing, in which the driver might be incapable to respond to the sudden takeover of control at all. The possible approaches to alleviate the startle during the takeover of control (i.e., the automation startle) include an offset- (i.e., either early- or delayed-), gradual yielding the controls to the driver. In the cases considered above, however, these approaches are hardly applicable because of (i) the presumed unpredictability of the events that result in the need of takeover of control, and (ii) the severe time constraints of the latter. Conversely, the objective of our research is to propose an approach of minimizing the need of yielding the control to the driver in challenging environmental conditions by guaranteeing an adequate automated control in these conditions. Focusing on slippery roads as an instance of challenging conditions, and steering control as an instance of control, we aim at developing such an automated steering that controls the car adequately in various road surfaces featuring low friction coefficients without the need of driver’s intervention.In order to develop such an automated steering we employed an in-house evolutionary computation framework – XML-based genetic programming (XGP) – which offers a flexible, portable, and human readable representation of the evolved optimal steering functions. The trial runs of the evolved steering functions were performed in the Open Source Racing Car Simulator (TORCS), which features a realistic, yet computationally efficient simulation of the car and its environment. The obtained experimental results indicate that due to the challenging dynamics of the unstable car on slippery roads, neither the canonical (tuned) servo-control (as a variant of PD) nor the (tuned) PID-controller could control the car adequately on slippery roads. On the other hand, the controller, featuring a relaxed, arbitrary structure evolved by XGP outperforms both the servo- and PID controllers in that it results in a minimal deviation of the car from its intended trajectory in rainy, snowy, and icy road conditions. Moreover, the evolved steering that employs anticipated perceptions is even superior as it could anticipate the imminent understeering of the car at the entry of the turns and consequently – to compensate for such an understeering by proactively turning the steering wheels in advance – well before entering the turn. The obtained results suggest a human competitiveness of the evolved automated steering as it outperforms the commonly used alternative steering controllers proposed by human experts. The research could be viewed as a step towards the evolutionary development of automated steering of cars in challenging environmental conditions. / 博士(工学) / Doctor of Philosophy in Engineering / 同志社大学 / Doshisha University

The introduction of autonomous vehicles in traffic is driven by expected gains in multiple areas, such as improvement of health and safety, better resource utilization, pollution reduction and greater convenience. The development of more competent algorithms will determine the rate and level of success for the ambitions around autonomous vehicles. In this thesis work an intersection management system for a mix of autonomous and manually driven vehicles is created. The purpose is to investigate the strategy to combine turn intention prediction for manually driven vehicles with scheduling of autonomous vehicle. The prediction method used is support vector machine (SVM) and scheduling of vehicles have been made by dividing the intersection into an occupancy grid and apply different safety levels. Real-life data comprising recordings of large volumes of traffic through an intersection has been combined with simulated vehicles to assess the relevance of the new algorithms. Measurements of collision rate and traffic flow showed that the algorithms behaved as expected. A miniature vehicle based on a prototype for an autonomous RC-car has been designed with the purpose of testing of the algorithms in a laboratory setting. / Införandet av autonoma fordon i trafiken drivs av förväntade vinster i flera områden, såsom förbättring av hälsa och säkerhet, bättre resursutnyttjande, minskning av föroreningar och ökad bekvämlighet. Utvecklingen av mer kompetenta algoritmer kommer att bestämma hastigheten och nivån på framgång för ambitionerna kring autonoma fordon. I detta examensarbete skapas ett korsningshanteringssystem för en blandning av autonoma och självkörande bilar. Syftet är att undersöka strategin att kombinera prediktion av hur manuellt styrda bilar kommer att svänga med att schemalägga autonoma bilar utifrån detta. Prediktionsmetoden som använts är support vector machine (SVM) och schemaläggning av bilar har gjorts genom att dela upp korsningen i ett occupancy grid och tillämpa olika säkerhetsmarginaler. Verklig data från inspelningar av stora volymer trafik genom en korsning har kombinerats med simulerade fordon för att bedöma relevansen av de nya algoritmerna. Mätningar av kollisioner och trafikflöde visade att algoritmerna uppträdde som förväntat. Ett miniatyrfordon baserat på en prototyp av en självkörande radiostyrd bil har tagits fram i syfte att testa algoritmerna i laboratoriemiljö.

For safe and reliable driving, it is essential that an autonomous vehicle can accurately perceive the surrounding environment. Modern sensor technologies used for perception, such as LiDAR and RADAR, deliver a large set of 3D measurement points known as a point cloud. There is a huge need to interpret the point cloud data to detect other road users, such as vehicles and pedestrians. Many research studies have proposed image-based models for 2D object detection. This thesis takes it a step further and aims to develop a LiDAR-based 3D object detection model that operates in real-time, with emphasis on autonomous driving scenarios. We propose 3D YOLO, an extension of YOLO (You Only Look Once), which is one of the fastest state-of-the-art 2D object detectors for images. The proposed model takes point cloud data as input and outputs 3D bounding boxes with class scores in real-time. Most of the existing 3D object detectors use hand-crafted features, while our model follows the end-to-end learning fashion, which removes manual feature engineering. 3D YOLO pipeline consists of two networks: (a) Feature Learning Network, an artificial neural network that transforms the input point cloud to a new feature space; (b) 3DNet, a novel convolutional neural network architecture based on YOLO that learns the shape description of the objects. Our experiments on the KITTI dataset shows that the 3D YOLO has high accuracy and outperforms the state-of-the-art LiDAR-based models in efficiency. This makes it a suitable candidate for deployment in autonomous vehicles. / För att autonoma fordon ska ha en god uppfattning av sin omgivning används moderna sensorer som LiDAR och RADAR. Dessa genererar en stor mängd 3-dimensionella datapunkter som kallas point clouds. Inom utvecklingen av autonoma fordon finns det ett stort behov av att tolka LiDAR-data samt klassificera medtrafikanter. Ett stort antal studier har gjorts om 2D-objektdetektering som analyserar bilder för att upptäcka fordon, men vi är intresserade av 3D-objektdetektering med hjälp av endast LiDAR data. Därför introducerar vi modellen 3D YOLO, som bygger på YOLO (You Only Look Once), som är en av de snabbaste state-of-the-art modellerna inom 2D-objektdetektering för bilder. 3D YOLO tar in ett point cloud och producerar 3D lådor som markerar de olika objekten samt anger objektets kategori. Vi har tränat och evaluerat modellen med den publika träningsdatan KITTI. Våra resultat visar att 3D YOLO är snabbare än dagens state-of-the-art LiDAR-baserade modeller med en hög träffsäkerhet. Detta gör den till en god kandidat för kunna användas av autonoma fordon.

This thesis tackles the problem of predicting the collision risk for vehicles driving in complex traffic scenes for a few seconds into the future. The method is based on previous research using dynamic Bayesian networks to represent the state of the system. Common risk prediction methods are often categorized into three different groups depending on their abstraction level. The most complex of these are interaction-aware models which take driver interactions into account. These models often suffer from high computational complexity which is a key limitation in practical use. The model studied in this work takes interactions between drivers into account by considering driver intentions and the traffic rules in the scene. The state of the traffic scene used in the model contains the physical state of vehicles, the intentions of drivers and the expected behaviour of drivers according to the traffic rules. To allow for real-time risk assessment, an approximate inference of the state given the noisy sensor measurements is done using sequential importance resampling. Two different measures of risk are studied. The first is based on driver intentions not matching the expected maneuver, which in turn could lead to a dangerous situation. The second measure is based on a trajectory prediction step and uses the two measures time to collision (TTC) and time to critical collision probability (TTCCP). The implemented model can be applied in complex traffic scenarios with numerous participants. In this work, we focus on intersection and roundabout scenarios. The model is tested on simulated and real data from these scenarios. %Simulations of these scenarios is used to test the model. In these qualitative tests, the model was able to correctly identify collisions a few seconds before they occur and is also able to avoid false positives by detecting the vehicles that will give way. / Detta arbete behandlar problemet att förutsäga kollisionsrisken för fordon som kör i komplexa trafikscenarier för några sekunder i framtiden. Metoden är baserad på tidigare forskning där dynamiska Bayesianska nätverk används för att representera systemets tillstånd. Vanliga riskprognosmetoder kategoriseras ofta i tre olika grupper beroende på deras abstraktionsnivå. De mest komplexa av dessa är interaktionsmedvetna modeller som tar hänsyn till förarnas interaktioner. Dessa modeller lider ofta av hög beräkningskomplexitet, vilket är en svår begränsning när det kommer till praktisk användning. Modellen som studeras i detta arbete tar hänsyn till interaktioner mellan förare genom att beakta förarnas avsikter och trafikreglerna i scenen. Tillståndet i trafikscenen som används i modellen innehåller fordonets fysiska tillstånd, förarnas avsikter och förarnas förväntade beteende enligt trafikreglerna. För att möjliggöra riskbedömning i realtid görs en approximativ inferens av tillståndet givet den brusiga sensordatan med hjälp av sekventiell vägd simulering. Två olika mått på risk studeras. Det första är baserat på förarnas avsikter, närmare bestämt att ta reda på om de inte överensstämmer med den förväntade manövern, vilket då skulle kunna leda till en farlig situation. Det andra riskmåttet är baserat på ett prediktionssteg som använder sig av time to collision (TTC) och time to critical collision probability (TTCCP). Den implementerade modellen kan tillämpas i komplexa trafikscenarier med många fordon. I detta arbete fokuserar vi på scerarier i korsningar och rondeller. Modellen testas på simulerad och verklig data från dessa scenarier. I dessa kvalitativa tester kunde modellen korrekt identifiera kollisioner några få sekunder innan de inträffade. Den kunde också undvika falsklarm genom att lista ut vilka fordon som kommer att lämna företräde.

What are pedestrian attitudes towards driverless vehicles that have no human driver? In this paper, we use virtual reality to simulate a virtual scene where pedestrians interact with driverless vehicles. This was an exploratory study where 15 users encounter a driverless vehicle at a crosswalk in the virtual scene. Data was collected in the form of video and audio recordings, semi-structured interviews and participant sketches to explain the crosswalk scenes they experience. An interaction design framework for vehicle-pedestrian interaction in an autonomous vehicle has been suggested which can be used to design and model driverless vehicle behaviour before the autonomous vehicle technology is deployed widely. / Vad är fotgängares inställning till förare utan fordon som inte har någon mänsklig förare? I det här dokumentet använder vi virtuell verklighet för att simulera en virtuell scen där fotgängare interagerar med förare utan bil. Det här var en undersökande studie där 15 användare möter ett förarefritt fordon vid en korsning i den virtuella scenen. Uppgifterna samlades i form av video- och ljudinspelningar, halvstrukturerade intervjuer och deltagarskisser för att förklara de övergripande scenerna de upplever. En ram för interaktionsdesign för fordonets fotgängarinteraktion i ett autonomt fordon har föreslagits, vilket kan användas för att utforma och modellera körlösa fordonsbeteenden innan den autonoma fordonstekniken används brett.