Advanced Driver Assistance Systems (ADAS) play a crucial role in enhancing road safety and driver comfort by providing real-time alerts and support. However, the effectiveness of ADAS signals in influencing driver situational awareness (SA) remains an area requiring further investigation. This study examines how different ADAS speed limit alert designs impact the three levels of SA: perception, comprehension, and projection.

A real-world driving experiment was conducted in Amersfoort along a 4.3 km route, comparing two ADAS systems with distinct signal designs. System A provided both the current and upcoming speed limits, while System B displayed only the current limit. Data collection included eye-tracking for SA1 (perception), self-reported questionnaires for SA2 (comprehension), and speed analysis for SA3 (projection). Statistical analyses, including significance tests, correlation analysis, and Linear Mixed Models (LMM), were employed to evaluate the influence of ADAS design and driver-related factors on SA.

Results indicate that while most SA indicators did not show significant differences between the two ADAS systems, System A consistently demonstrated advantages in visual attention allocation and comprehension. Additionally, driver experience and familiarity with ADAS were found to influence SA, with experienced drivers exhibiting reduced cognitive load and faster responses. A weak correlation between different SA levels suggests that drivers integrate ADAS alerts with external environmental cues.

This study highlights the importance of designing ADAS alerts that provide anticipatory information to enhance driver awareness and decision-making. Future research should expand sample sizes, incorporate additional SA indicators, and explore the long-term effects of ADAS exposure. These findings offer valuable insights for optimizing ADAS signal design to improve road safety and driving performance.

Background
The steady development of automated vehicles (AVs) promises significant benefits in terms of traffic safety and efficiency. However, the transition to fully AVs and their deployment on the road will be gradual, leading to a phase of mixed-traffic conditions where AVs at various levels coexist with human-driven vehicles (HDVs). This transition poses unprecedented hurdles, requiring a deeper understanding of the emerging challenges for AVs in sensing and perceiving road environments, as well as in the novel interactions between AVs and HDVs. Furthermore, the social compliance of AVs and the optimization of their deployment strategies need to be considered as well.

Contents of this Thesis
This thesis addresses the multifaceted challenges associated with AVs’ development and deployment in mixed-traffic environments. The main objective of this thesis is to enhance the capabilities of AVs enabling them with a wider Operational Design Domain (ODD) and thus facilitate the implementation of safe, efficient, and socially compliant automated driving in mixed traffic. Referring to the modular design of AV systems, three key perspectives, i.e., sensing and perception, anomaly detection, as well as planning and control, are tackled in this thesis. To be specific:
Chapters 2-4 focus on enhancing sensing and perception capabilities through the development of hybrid spatial-temporal deep learning models and self-supervised pretraining methods. Lane detection is chosen as the focus of these chapters since it is vital for current vehicle localization and positioning, and it is also the foundation of various automated driving features. The main findings of these chapters are summarized as follows.
Chapter 2 presents a pioneering hybrid spatial-temporal sequence-to-one deep learning architecture tailored for vision-based lane detection tasks. By integrating the spatial convolutional neural network (SCNN) with spatial-temporal Recurrent Neural Network (RNN) modules, this architecture effectively captures correlations and dependencies among continuous image frames. Through extensive experimentation on various driving scenes, including challenging scenarios, the proposed model variants exhibit superior performance over existing state-of-the-art models. Notably, even the lighter model variants demonstrate remarkable accuracy, outperforming their counterparts while maintaining lower computational complexity.
Building upon the foundation laid in Chapter 2, Chapter 3 focuses on refining vision-based sensing and perception through the development of customized spatial-temporal attention mechanisms. These mechanisms, including temporal attention, spatial-temporal attention, and spatial-temporal attention with fully connected layers, are meticulously designed to optimize the utilization of spatial-temporal correlations across different regions of interest within the consecutive image frames. Leveraging linear Long Short Term Memory (LSTM) neural networks in conjunction with the proposed attention blocks, this chapter demonstrates the feasibility of lightweight and computationally efficient solutions for sequential deep neural networks (DNNs). Through rigorous experimentation, ablation studies, and comparative analysis across diverse datasets, the effectiveness of the proposed attention mechanisms in enhancing lane detection performance is convincingly established.
In Chapter 4, the exploration of enhancing vision-based sensing and perception capabilities continues with the introduction of a self-supervised pretraining method employing masked sequential autoencoders (MSAE). This innovative approach leverages both labelled and unlabelled data to improve detection accuracy and expedite the training process of DNN models dedicated to lane detection tasks. Additionally, a customized Focal Loss based PolyLoss is introduced to further enhance the detection accuracy. Through comprehensive experimentation and comparative analysis, the efficacy of the proposed pretraining method and loss function is demonstrated, showcasing substantial improvements in lane detection performance across diverse driving scenarios. Specifically, the utilization of MSAE-based pretraining and the adoption of the customized PolyLoss result in superior performance metrics, underscoring the pivotal role of self-supervised learning techniques and tailored loss functions in fortifying the robustness and efficiency of vision-based sensing and perception systems in AVs.
These chapters address the challenges of vision-based lane detection, crucial for AV navigation and safety.

Chapters 5-6 delve into anomaly detection, investigating techniques for identifying abnormal lane rendering in digital map applications and detecting anomalies in driving behaviour.
Chapter 5 introduces an innovative approach to anomaly detection in lane rendering images of digital map applications, utilizing Transformer-based models with self-supervised pretraining and customized fine-tuning. By transforming anomaly detection into a classification problem, the chapter proposes a four-phase pipeline that includes data pre-processing, self-supervised pre-training with masked image modelling (MiM), customized fine-tuning using cross-entropy-based loss, and post-processing. Experimental results demonstrate the pipeline’s effectiveness, with significant improvements in detection accuracy and reduced training time achieved through self-supervised pre-training. Ablation studies regarding tackling the problem with different numbers of classes further validate the pipeline’s performance enhancements, particularly in addressing data imbalance. This approach not only enhances anomaly detection accuracy but also contributes to reducing labour costs associated with manual labelling and anomaly detection efforts, offering significant societal benefits.
Additionally, Chapter 6 explores the critical task of detecting abnormal driving behaviour, addressing the need for more feasible and efficient approaches by leveraging semi-supervised ML methods. Utilizing large-scale real-world driving data, the study develops a semi-supervised ML model based on Hierarchical Extreme Learning Machines (HELM). This approach utilizes partly labelled data and introduces Surrogate Safety Measures (SSMs) (specifically the event-baed safety indicators of Two-Dimensional Time-To-Collision (2D-TTC)) as the pivotal input features to enhance performance. Results demonstrate the effectiveness of the proposed semi-supervised ML model, showcasing superior performance compared to baseline methods. The integration of SSMs significantly improves detection accuracy, highlighting their significant role in enhancing model performance. By leveraging unlabelled data for training and only a small sample of labelled data for fine-tuning, the proposed semi-supervised approach achieves competitive performance while reducing dependency on fully labelled datasets, making it suitable for real-world applications.
To sum up, the exploration of semi-supervised and self-supervised ML methods presents promising avenues in anomaly detection. The pioneering research presented in this thesis represents a significant stride towards leveraging data-driven ML-based anomaly detection methodologies to enhance the safety of driving.

Chapters 7-9 shift the focus to planning and control strategies for AVs, presenting a comprehensive examination of decision-making frameworks and control algorithms. These chapters introduce a conceptual framework aimed at fostering socially compliant driving behaviour and propose a range of model-based and learning-based approaches.
Chapter 7 lays the groundwork by introducing a conceptual framework that emphasizes socially compliant automated driving. This framework encompasses various social components such as cultural nuances, norms, and driving styles. A key innovation is the introduction of bidirectional behavioural adaptation, highlighting the dynamic interactions between AVs and human drivers. Furthermore, the framework advocates for the incorporation of a spatial-temporal memory module to enable continuous refinement of driving strategies, thereby promoting adaptability and safety in diverse traffic scenarios. Validation through an online expert survey lends credence to the framework’s efficacy. This conceptual framework lays a solid foundation for learning-based and model-based approaches for implementing planning and control algorithms for automated driving.
In the learning-based approach explored in Chapter 8, Deep Reinforcement Learning (DRL) takes centre stage, with a focus on integrating safety, efficiency, comfort level, and energy consumption considerations into the learning framework. Multiple DRL algorithms are evaluated across diverse driving manoeuvres, particularly roundabout driving, highlighting the importance of real-world requirements in reward function design and simulation-based training. Among the compared DRL algorithms, Trust Region Policy Optimization (TRPO) emerges as leading in safety and efficiency, while Proximal Policy Optimization (PPO) excels in comfort during roundabout driving. Moreover, the extension of the training environment to encompass various driving scenarios showcases the adaptability of DRL models to train a uniform driving model for real traffic environments, signalling promising avenues for future research.
Regarding the model-based approach, Chapter 9 introduces the DRF-SVO-MPCC algorithm, aimed at enhancing AVs’ understandability and predictability to human drivers, particularly during interactions with HDVs when driving through the roundabouts, as this challenging manoeuvre involves large curvature and tackles both longitudinal and lateral control. This algorithm integrates the perceived Driving Risk Field (DRF), Social Value Orientation (SVO), and Model Predictive Contouring Control (MPCC), enabling AVs to navigate social scenarios with sensitivity to the welfare of surrounding HDVs. Simulation experiments, conducted on various roundabout scenarios, underscore the algorithm’s superiority in trajectory tracking and adaptability to different driving styles, ensuring safety and social compliance. The findings illuminate the potential of the DRF-SVO-MPCC algorithm in fostering harmonious interactions between AVs and HDVs, setting a precedent for socially aware automated driving systems.

Overall, this thesis represents a solid endeavour to advance the planning and control capabilities of AVs in mixed-traffic environments. Through the development of novel conceptual frameworks and innovative model-based and learning-based algorithmic solutions, it lays the groundwork for the realization of safe, efficient, socially compliant, and adaptable automated driving, contributing to safer and more harmonious transportation systems.

Conclusion and perspectives
In summary, this thesis contributes to advancing the knowledge of how to improve automated driving systems in the realms of sensing and perception, anomaly detection, as well as planning and control. By integrating theoretical frameworks, methodological innovations, and data-driven empirical evaluations, notable progress has been achieved in fostering the development of safe, efficient, and socially compliant automated driving within mixed-traffic environments.
Despite the considerable progress made, several directions for future research have been identified. These include the imperative for more expansive high-quality datasets, exploration of domain adaptation techniques for both sensing and anomaly detection tasks, as well as the seamless integration of model-based and learning-based methodologies for planning and control. Additionally, transitioning towards a unified driving model and effectively addressing the complexities of multi-agent interactions in intricate urban settings remain pivotal areas for further exploration. Furthermore, interdisciplinary collaboration will be instrumental in harnessing the full potential of automated vehicles to revolutionize transportation systems.

As automated vehicles (AVs) become more common, their influence on human-driven vehicles (HDVs) in mixed traffic is increasingly relevant. This dissertation explores how AVs affect HDV driving behavior—specifically car-following, overtaking, and gap acceptance—and how these behavioral adaptations influence traffic efficiency. Combining driving simulator experiments, field tests, and traffic microsimulation, the research provides insights into the dynamics of mixed traffic and their implications for infrastructure and AV deployment.

This study examines the impact of Adaptive Cruise Control (ACC) systems on traffic flow efficiency and road safety, with a particular emphasis on transition control between automated and manual driving. Empirical data from the OpenACC database and the SAE Level 2 naturalistic driving study were analyzed to capture key ACC characteristics, which were then used in simulations to reflect the variability among different ACC systems. Simulations were conducted on a section of the Dutch A13 highway, with ACC market penetration rates (MPRs) set at 25% and 75%. The ACC and transition control models were integrated as external driver models in PTV VISSIM under the assumption of nearly ideal human driver responses, excluding potential response delays and the string instability typically associated with ACC systems.

The results indicate that ACC systems with transition control models significantly enhance traffic flow efficiency and safety, particularly at higher MPRs. The introduction of ACC vehicles reduced congestion and increased time-to-collision (TTC) values, reflecting improved traffic safety. However, the study highlights limitations, including simplified human driving behaviors and the exclusion of string stability effects. Future research should focus on more complex driving scenarios, such as urban environments, and enhance data collection methods to further understand ACC system performance in diverse traffic conditions.

Past bicyclist subjective safety studies have evaluated different intersection types and bicycle facilities at roundabouts. None have focused on the different design elements of a dedicated bicycle facility and analyzed both perceptions of safety and comfort of bicyclists in a single model. To address this, this study focuses on the subjective safety and comfort perceptions of bicyclists at Dutch urban roundabouts, focusing on certain bicycle design elements. Infrastructure evaluated included: bi-directional bicycle paths; yielding priority; bicycle volumes; the presence of art or advertisements in the center island; buffer width between the bicycle path and the vehicle lane;
and the number of vehicular entrance/exit points. Using a multivariate random effects ordered probit model, the results revealed that roundabouts where cars have to yield to bicycles and fewer vehicular legs are perceived as safer and more comfortable by cyclists. Conversely, high bicycle volumes and increasing age negatively impacted perceptions of safety. The study found that bicyclists’ place of residence, their likelihood to commit traffic violations, their recent crash history, and the type of bicycle used (e.g., cargo bicycle, e-bicycle, etc.) significantly affect their perceptions of safety. The results suggest that consistent adherence to Dutch design guidelines can enhance bicyclists’ sense of comfort and safety at roundabouts. The findings highlight the need for designers to consider bicyclists’ perspectives in roundabout design. To improve perceived safety and comfort for bicyclists recommendations include: ensuring bicycle yielding priority; designing roundabouts to reduce bicycle congestion to the extent possible; and maintaining uniformity in bicycle infrastructure.

The introduction of Highly Automated Vehicles (HAVs) presents challenges for everyday transport, particularly impacting cyclists, who are part of what is called vulnerable road users. While HAVs promise to optimise traffic flow and enhance safety, concerns about their safety remain. This research addresses the gap in understanding cyclists' expectations and perceptions of social driving behaviour in The Netherlands. By developing a conceptual framework and a questionnaire, the study explores the interactions between cyclists and HAVs, focusing on trust, perceived safety and perceived social behaviour of cyclists in urban settings like The Hague. Findings reveal that cyclists' comfortability with sharing the roads with HAVs and their belief that HAVs are more likely to follow the rules influence their opinions. Additionally, the use of electronic Human-Machine Interface (eHMI) in HAVs, seems to positively affect cyclists’ trust, perceived safety and perceived social behaviour. Notably, it is found that demographic factors have little influence on cyclists' attitudes towards HAVs. The study underscores the importance of predictability in fostering higher safety ratings and demonstrates how the characteristics and driving behaviour of HAVs affect cyclist perceptions. These insights can guide policymakers and developers in devising solutions that enhance the safety and well-being of all road users, ensuring the effective integration of HAV technology while minimising potential drawbacks.

In the Netherlands, roads are classified into three categories: access roads, distributor roads, and through roads, each with distinct functions and design criteria. Nonetheless, within built-up areas, some roads have characteristics of both access and distributor roads, leading to inadequate separation of various traffic flows while having a speed limit of 50 km/h resulting in a heightened risk of accidents. These are so-called 'grey roads'. To improve road safety on grey roads, recent adjustments have been made to the design requirements. Whereas the standard speed limit on distributor roads was previously 50 km/h, it is now 30 km/h unless 50 km/h can be safely implemented. The requirements for when a road is safe enough for a 50 km/h speed limit have been tightened. Nevertheless, a combination of parallel parking and a speed limit of 50 km/h are still permitted in these tightened design requirements. Literature indicates that roads with parallel parked vehicles have an increased risk of accidents and that higher speeds increase the impact of an accident.

Due to the recent implementation of these design requirement adjustments, there is insufficient data on road accidents to determine whether road safety would improve by lowering the speed limit to 30 km/h on roads with parallel parking. Therefore, a surrogate safety measure was necessary. In this research, situational workload was chosen as the surrogate safety measure. This thesis investigated the following research question: "What is the effect of parallel parking in combination with speed reduction from 50 km/h to 30 km/h on the driver's situational workload, the feeling of safety and credibility of the speed?".

Situational workload was derived from the concept of 'mental workload,' which refers to the dynamic relationship between a driver's capabilities and the task demand. A driver maintains control over the vehicle and adequate situational awareness if the task demand is within their capabilities. If the task demand exceeds the capability, the situational awareness is endangered and the risk of making errors increases. Changes in mental workload are often assessed using primary or secondary task performance measures: physiological measures, or subjective reflections. This research does not capture the full task demand or the complete capabilities of the drivers, since virtual reality (VR) was used as a research methodology. Consequently, it is unclear what each driver's baseline mental workload is, how it is affected by parallel parking and speed, and whether the increase in mental workload endangers driving performance.
To address this, the term situational workload was introduced in this study. Situational workload captures the influence of road conditions on self-estimated driving performance, self-reported situational awareness, and self-reported workload, based on a combination of subjective reflections and primary task performance measures. The self-estimated driving performance was assessed by evaluating the collision avoidance estimation and the situational awareness was measured by the level of overview on the road. The self-reported workload was complemented by the self-reported level of attention required for the driving task to measure the difficulty of the driving task...

Cycling as a mode of transport is getting increasingly more attention from academics and governments around the world. It improves the physical and mental well being of citizens, it is efficient in terms of space and it provides economic benefits. However, traffic engineers and urban planners of these cities currently lack the means to quantitatively test their bicycle infrastructure designs. Simulation models have become an important tool for the transport planner as it is cheap and effective. But simulation models of cyclist behaviour are an understudied subject.

This is not only an issue for cities that attempt to transform there mobility landscape to a more bicycle-friendly one, but also for established cycling cities that face new challenges because of the introduction of new vehicle types (e.g., e-bike, e-step) and new mobility business models (e.g., shared mobility, flash delivery). The academic world is also affected by the fact that cyclist simulation is understudied. The literature review of this thesis reveals that the academic landscape on this subject is foggy and unclear. The most popular model in academia is the Social Force Model (SFM). However, just the SFM leads to unsatisfactory behaviour, as is confirmed in the case study of this thesis. Thus, the academics propose a hybrid model. Herein, the SFM is constrained by additional rules, regimes or decision making processes. However, these propositions are tailored to cycling in a specific context and they are created in unmentioned software, using unmentioned programming languages. This hampers the possibility to replicate and verify their study.

The primary conclusion is that academics and simulation software developers should work towards a standardised behavioural model for cyclists, similar to the SFM for pedestrians and the car-following model for motorised traffic. An example of a leading principle for cycling behaviour could be a model where cyclists move along predetermined trajectories rather than being led by a force. This seems to solve the challenges researchers face regarding the operational behaviour of the cyclists.
It is also advised in future research to discuss the semantics of cyclist behaviour simulation. Literature that used this principle of predetermined trajectories still referred to their model as a model based on the SFM. This understates the fundamental criticism of these researchers on the SFM.

Lastly, commercial software companies are advised to actively partake in this process of creating a standardised model for cyclist simulation. With pro-cycling policy becoming more popular in cities all over the world, it is expected that municipalities and research institutes will show interest in a standardised model for cyclist behaviour when offered by a simulation software company. Moreover, such a standardised model incorporated in an applied simulation software should improve the verifiability and replicability of academic research. As a properly calibrated model for cyclist behaviour, would mean that less academics are inclined to create their own custom model for their research.

If these proposed developments come to fruition, other research subjects and applications become possible or more accessible. It should allow peer reviewers to test a researcher’s simulation model more easily. It enables urban planners and traffic engineers of local governments to quantitatively test their infrastructure designs and it enables academics to study the rapidly changing urban mobility landscape. For instance, it provides a possibility to experiment with infrastructure designs, aiming to tackle the contemporary challenges of established cycling cities.

Electric vehicles (EVs) are required according to the EU legislation from 2014 to install Acoustic Vehicle Alerting Systems (AVAS) to convey information about the EVs through sound signals to other road users, ensuring their safety. Meanwhile,autonomous vehicles (AVs) are also recommended to be equipped with an External Human-Machine Interface (eHMI) system, enabling better interaction with other road users. . However, for the special group of autonomous electric vehicles (AEVs), considering that equipping both eHMI and AVAS, given their overlapping functionality in conveying information, might result in a waste of resources, it remains unclear whether it's feasible to use only eHMI for communication with road users as a cost-saving measure, thereby omitting AVAS or perhaps adding only a simple sound signal to replace AVAS. This study aims to explore whether autonomous electric vehicles need to add extra noise to provide more information to other vulnerable road users, such as cyclists and pedestrians. In this study, we will use Virtual Reality (VR) technology for simulation experiments. Utilizing VR, as opposed to real-world experiments, allows for precise control over the variables in each experiment, ensuring that experiments can be conducted under the same conditions multiple times, thus enhancing reliability. On the other hand, VR can ensure the safety of experiment participants while providing them with an immersive experience, especially in experiments related to traffic safety research.

In this experiment, we will utilize the Unreal Engine to create a simulated testing environment, focusing on observing participants' (acting as cyclists) reactions to a series of variables. These include different environmental noises, the warning distance of sound signals emitted by autonomous vehicles, and the autonomous vehicles themselves. The sound signal warning distance specifically refers to a mechanism where, when pedestrians or cyclists enter a predefined range of the vehicle, it automatically emits a warning signal. This signal is to alert the approaching individuals that the vehicle is in a ready state and may proceed to the next action, thereby enhancing safety interaction and awareness on the road.

The data collected from this experiment will be divided into two parts: first, the participants' reaction times, speed adjustments, and distances obtained from VR devices; second, their perceptions of the trust, perceived safety, and comfort of interactions with autonomous electric vehicles gathered through post-experiment surveys.

The results of this experiment will assist policymakers in refining relevant laws and regulations, as the existing legislation concerning electric vehicles and AVAS do not take autonomous vehicles into account. It will also provide theoretical support for car manufacturers in the design of autonomous electric vehicles.

This study investigates how Dutch municipalities can measure and use subjective road safety. A literature review was conducted, analyzing 31 scientific sources. In addition, gray literature and practical differences were examined, involving 13 different parties. Interviews were also held with six municipalities to discuss their perspectives on the definition of subjective road safety, the data they currently have, their goals regarding subjective road safety, the resources available for further measurements, and their most pressing needs. The study found that there is still significant ambiguity within municipalities regarding the definition of subjective road safety. For many municipalities, the individual’s perception of safety plays a central role. Furthermore, there is confusion about the relationship between objective and subjective road safety. Municipalities need more concrete goals to make informed decisions about whether to measure subjective road safety reactively or proactively. Policies can be developed based on these goals, but it is crucial that objective road safety is not compromised. Using subjective road safety as a measure to gain insight into objective road safety is not advisable, as the relationship between these two dimensions is not yet fully understood. While they coexist and influence each other, they function independently.

Connected and Automated Vehicles (CAVs) have the poten- tial to revolutionize transportation systems, but their integration with human-driven vehicles (HDVs) in mixed traffic environments presents significant challenges, particularly in complex scenarios such as on-ramp merging. This paper addresses the challenge of on-ramp merging for CAVs in mixed traffic environments, proposing a novel approach called QMIX-QLambdaM. We formulate the problem as a Centralized Train- ing with Decentralized Execution (CTDE) Cooperative Multi-Agent Re- inforcement Learning (MARL) task, capable of handling dynamic sce- narios with both CAVs and HDVs. QMIX-QLambdaM enhances the QMIX algorithm by incorporating Q(λ) returns for improved value es- timation and an action masking mechanism for safer action selection. Our comprehensive experiments demonstrate that QMIX-QLambdaM consistently outperforms state-of-the-art algorithms, including QMIX, MAA2C, and COMA, across various performance metrics related to traf- fic efficiency and safety. The proposed method exhibits superior adapt- ability across different traffic densities, maintaining high performance in terms of safety, efficiency, and overall rewards. Furthermore, case stud- ies illustrate QMIX-QLambdaM’s ability to generate effective strate- gic control for both main-lane and merging-lane vehicles, showcasing smoother driving behavior and better collision avoidance compared to baseline methods. The learning curve comparison also reveals QMIX- QLambdaM’s advantage in credit assignment compared to other CTDE baselines for the formulated problem. The code are available at https: //github.com/ayton-zhang/MARL_qmix_merging.

In order to increase traffic safety in the Netherlands, the government implemented the GOW30-policy. However, evaluation of this policy shows that drivers do not comply with this newly imposed speed limit sufficiently. For this reason, roads need to be designed in such a way that speed limits are perceived as credible. At the same time, the need for sustainable road designs arises, due to factors as increased urban heat stress and population growth. Thus, road designers need to make trade-offs between spatial quality and mobility. Incorporating greenery is a promising way to reduce urban heat stress, but the presence of greenery, particularly thick trees, can also increase the severity of traffic accidents. This study evaluated the impact of roadside greenery on driving behaviour, utilising a driving simulator experiment complemented by a post-experiment survey. Linear Mixed Model analyses (LLM’s) together with survey insights revealed that perceived driving comfort leads to increased driving speeds. Comfort increases on wider roads and on roads with more overview. For this reason, it is not recommended to incorporate low greenery on GOW30 roads. Moreover, high and dense greenery is shown to reduce driving speeds while promoting better positioning on the road. It is furthermore recommended that research is performed in real-life, whilst evaluating roadside greenery in combination with cycling lanes under varying traffic and weather conditions.

This dissertation explores what road characteristics trigger drivers’ speed adjustments when approaching freeway curves. It combines speed prediction modelling and human factors research methods. The results show that drivers primarily consider visible cues such as the preceding roadway, deflection angle, and the number of lanes, as opposed to traditional factors like horizontal radius or speed signs, when starting to decelerate. The study advocates for integrating driver perspectives into road design.

This research delves into the transformative potential of self-driving vehicles by investigating their impact on passengers’ happiness. As autonomous transportation technology rapidly evolves, understanding the user experience within these vehicles becomes essential. To investigate the happiness of the self-driving vehicles’ passengers, which is defined as positive emotions and cognitive well-being during self-driving rides, real-traffic test rides were conducted among 31 participants, companied with two before-ride and after-ride surveys. The study employs a comprehensive approach, combining self-reported survey data and biometric measurements, which includes the participants’ heart rate and the eye movement, to investigate passengers' happiness. The results present an overall positive emotions and positive attitudes towards the self-driving vehicle. Moreover, the findings present a notable shift in passengers' attitudes, with originally neutral sentiments transitioning to positive perceptions following the test ride. Participants exhibited various activities during the ride, enhanced comfort with the concept, and an improved satisfaction with self-driving technology. Remarkably, statistical trends suggest that self-driving vehicles hold the potential to alleviate stress and optimize time management, positively impacting passengers' overall well-being. Furthermore, biometric data of the participants presented participants’ different physical reaction on different traffic scenarios and indicated the happiness and well-being of the participant. The research emphasizes the broad implications of self-driving technology on individual happiness, concerning both emotions and attitudes, extending beyond functional enhancements to encompass passenger happiness and societal integration.

Lane-changing behaviour detection is a critical aspect of driving safety and traffic management. This study focuses on detecting sudden lane changes as a subset of abnormal driving behaviours. By analyzing the characteristics of abrupt lane changes, the aim is to develop effective data-driven unsupervised machine learning (ML) methods for their detection and classification. Three unsupervised ML models, namely Isolation Forest, Local Outlier Factor, and Robust Covariance are evaluated and compared using a dataset of lane-change events. The results show that the Isolation Forest and Local Outlier Factor models outperform the Robust Covariance model, with the Local Outlier Factor model excelling in precision and overall accuracy, achieving the best overall detection rate. Both Robust Covariance and Isolation Forest deliver satisfactory results. Conversely, the Robust Covariance model exhibits poor performance. The findings verify the capability of data-driven ML methods for enhancing road safety and driving experiences through effective detection of sudden lane changes using vehicle motion information data. Future work involves further improving the accuracy and reliability of the ML models, validating their generalizability on larger datasets, incorporating contextual information, and exploring their real-time implementation in driving assistance systems.

Autonomous vehicles (AVs) can solve a lot of problems related to traffic safety, comfort and congestion. While the sensors used by these vehicles are getting cheaper, more accurate, and software is improving, the first completely automated vehicle does not exist yet. When AVs start gradually driving on the roads, there will be a mixed traffic situation consisting of AVs and human driven vehicles. To understand the implications on the traffic performance, this thesis will implement a Model Predictive Controller (MPC) and the behavioural models for AVs into a numerical simulation in SUMO. The MPC can be used to control multiple AVs in larger traffic simulations to analyse the effects of certain penetration rates and traffic densities when changing the AVs driving behaviour. Both the Post-Encroachment Time (PET) and Time To Collision (TTC) will be used for decision making in overtaking and collision avoidance strategies. In this thesis, the MPC which is solving a quadratic programming problem has proven to successfully overtake other vehicles in an on- and off-ramp highway scenario which results in a smaller average travel time for the AVs than for the human drivers. It is shown that the decrease in average travel time for the AVs was not at the cost of the Human Vehicles (HVs). This is done by decreasing the TTC used by the AVs as a headway for longitudinal overtaking behaviour, and the PET used for lateral lane changing behaviour, independently to a minimum of 0.1 seconds to guaranty safety, and measure the average travel times for both HVs and AVs. Multiple simulations using different penetration rates ranging from 0 to 40 percent also show that the decrease in average travel time for the AVs was not at the cost of the HVs. We can conclude that using an MPC integrated together with the behavioural rules of human drivers into numerical simulation can give a good indication of the implications that AVs have on traffic performance.

This is a research project for course CIE5090-09. The research aims to investigate the impact of the light-mopeds on the cyclists’ perceived safety. An online questionnaire is delivered to collect the participants’ feelings on the perceived safety of the given scenarios, and the reasons why they are feeling unsafe, as well as their intended reaction to the light-mopeds. The results shows that the participants would in general feel unsafe when they are in the conflicting directions with the light-mopeds. The leading reason for them to feel unsafe is the turning and speeding of the light-mopeds. Moreover, the results shows that most of the participants would take evasive actions, for instance to give their way or slow down, when feeling unsafe interacting with the light-mopeds.

In the coming decade, Advanced Driver Assistance Systems (ADAS) will play a crucial role in improving traffic safety within the European Union (EU). Notably, the European Commission mandates that from July 2024, all new vehicles must be equipped with Lane Keeping Assistance (LKA) systems. LKA systems provide force feedback to drivers when vehicles cross lane markings, employing cameras, radar, or LiDAR systems for lane detection.

To optimize LKA performance, this research focuses on understanding the influence of lane marking properties and adverse scenarios on detection ability.

Field Test and Setup:
A field test was conducted at a track in Lelystad, featuring various state-of-the-art lane markings: an old white paint lane marking as a reference, and three new lane markings (two tape types from 3M and one cold spray plast from Triflex). Lane markings were tested under dry and wet conditions.

Data Collection:
Measurements included luminance coefficient (Qd) and retroreflectivity values (Rl for dry and Rw for wet) of lane markings and asphalt, along with contrast ratios. Test runs were conducted during sunset, in complete darkness, and under different scenarios involving oncoming traffic and street lights.

Analysis and Findings:
After analyzing 414 valid runs, key findings were identified:

Lane markings are 3.3 times more likely to be detected in dry versus wet conditions.
Driving towards a light source reduces detection likelihood by 4.5 to 5 times compared to driving away from the light.
Higher wet retroreflectivity (Rw) improves LKA system performance.
Oncoming traffic with main beam headlights reduces detection likelihood significantly.
Bright light sources negatively affect contrast in camera-detected images.
New lane markings outperform old white paint markings by 2.1 to 4.3 times in detection likelihood.

Implications and Recommendations:
The study confirms the importance of lane marking detectability for LKA systems reliant on cameras. Enhancing retroreflection and contrast improves detection ratios, thereby enhancing traffic safety and potentially preventing accidents. Further research is recommended on the impact of oncoming traffic and street lights, including light type and brightness.

For future lane marking developments, testing under adverse conditions with varying light sources and wet surfaces is advised to optimize LKA system performance and ensure safety on the roads.

Understanding human behavior in overtaking scenarios is crucial for enhancing road safety in mixed traffic with automated vehicles (AVs). Modeling plays a pivotal role in advancing our comprehension of human overtaking behavior in dynamically evolving scenarios. Currently, our understanding of overtaking behavior primarily revolves around straightforward interactions with human-driven vehicles (HDVs). To address this gap, we conducted a ``reverse" Wizard-of-Oz driving simulator experiment with 30 participants interacting with both oncoming AVs and HDVs, featuring time-varying dynamics. We hypothesized that the type of oncoming vehicle (AV or HDV) does not significantly influence gap acceptance during overtaking, while we anticipated an increase in gap acceptance when the oncoming vehicle briefly decelerates during interactions with the human ego-vehicle driver. Our findings reveal that participants did not significantly alter their overtaking behavior when interacting with oncoming AVs compared to HDVs. Surprisingly, brief decelerations in the oncoming vehicle's velocity did not significantly affect the decision-making processes of overtaking. Moreover, our results reinforced previous insights into the significance of the initial distance and time-to-arrival to the oncoming vehicle, and the ego-vehicle velocity on participants' overtaking behavior. We highlight the potential of simple drift-diffusion models (DDMs), a subset of cognitive models, in understanding human overtaking behavior in dynamically evolving scenarios involving oncoming AVs. Our proposed model accurately captures qualitative patterns in gap acceptance during these intricate overtaking scenarios, further advancing the ongoing development of safer interactions between human drivers and AVs during overtaking maneuvers.

E-bikes have gained global popularity due to their environmentally friendly and sustainable attributes, as well as their ability to provide fast speeds and power assistance. However, the increasing popularity of e-bikes has introduced conflicts and crashes with other road users, especially the interaction with conventional bikes, as e-bikes and bikes share the same cycling infrastructure. To study the perceived risk of interaction with e-bikes, this research delves into the various factors, in terms of traffic environment, bicycle type, and individual factors.
This research has used the video-based survey method to measure the perceived risk of cyclists and presented traffic environment and bicycle types in hypothetical traffic scenarios with pre-recorded videos. A questionnaire is designed to investigate participants` perceived risk of hypothetical bicycle interactions and their personal information on demographics, cycling experience, competence of riding skills, cycling behaviors, and expectancy on e-bikes. Moreover, the perceived risk is measured by two items separately: the rate of perceived risk and the likelihood of being involved in a crash. After recruitment and data collection, the effects of various factors have been estimated by the random-effects ordered logit model.
Results implied that compared to interaction between conventional bikes, riding on an e-bike when encountering a conventional bike decreased the perceived risk. Moreover, riding at peak hours and having a conflict with left-turning cyclists was found to increase the perceived risk, while riding at a large intersection was found to decrease the perceived risk. In terms of individual factors, the experience of bicycle crashes and the preference to use an e-bike were found positively related to the perceived risk. While the competence of cycling skills and age were found negatively associated with perceived risk. Additionally, it was found that traffic environment factors were more prominent in predicting the rate of perceived risk, while individual factors had a stronger influence on predicting the likelihood of being involved in a crash.
These findings provide insights into the influence of e-bikes in shared cycling spaces, underscore the importance of traffic management and safety measures in crowded bicycle traffic, and emphasize the significance of intervention and educational initiatives for cyclists to decrease the perceived risk and enhance cycling safety.