Human drivers routinely make interaction decisions by integrating multimodal, noisy perceptual information over time to decide whether to proceed or yield. Understanding the cognitive mechanisms underlying such decisions is crucial for explaining driver behaviour and for developing models that generalise across traffic scenarios.
This study examines human driver decision-making in a narrow-passage gap-acceptance task, an interaction scenario characterised by ambiguity of priority and described by interdependent kinematic and visual information. Behavioural data from laboratory ($N=36$) and ($N=175$) online experiments were analysed to assess whether decision dynamics generalise across populations with differing variability and experimental control.
Across datasets, decision behaviour showed systematic dependencies on kinematic conditions, with increased reaction times (RTs) and choice variability in ambiguous situations. Longitudinal kinematics-based drift diffusion models (LK-DDMs) captured both decision proportions and reaction-time distributions in the narrow passage task and generalised across datasets with differing coverage of the experimental conditions, including lab and online data. The same accumulation framework transferred to a related decision-making task, indicating that the inferred dynamics are not task-specific. Incorporating visual looming in the drift function yielded selective improvements in short-distance Wait decisions, without global gains in reaction-time accuracy.
OSF link: https://osf.io/5d6em/overview?view_only=18580aa504f24d92a55cbccf53bb5deb ... Read more

Human driver models are essential for the development and testing of Automated Driving Systems (ADS), yet current approaches often struggle to capture the complex, stochastic nature of human tactical decision-making. Large Language Models (LLMs) have emerged as potential reasoning agents capable of emulating human-like social behaviour, but their application as direct vehicle control agents remains largely underexplored.

This thesis investigates the extent to which a base LLM, guided by systematic prompt engineering, can replicate the tactical decisions and control of human drivers in a 1-D highway merging scenario. Using the OpenAI o3 model, an LLM-driven agent was developed and systematically benchmarked against a dataset of human driver behaviour recorded in a simulator experiment.
The study utilised Linear Mixed-Effects Regression (LMER) to analyse decision-making mechanisms and performed a sensitivity analysis using the Google Gemini-2.5-pro model to assess generalisability.

The results demonstrate that the LLM agent successfully replicated high-level tactical behaviours, satisfying qualitative criteria such as symmetrical yielding in neutral conditions and increased yield rates when the opposing vehicle held a headway advantage. However, a fundamental disparity was observed in operational control. While human drivers relied significantly on relative velocity to negotiate merges (p = 1.88 × 10−26), the LLM adopted a conservative, calculation-heavy gap-based strategy driven by absolute distance, resulting in average safety margins more than double the human benchmark (9.18 m vs. 3.85 m). Furthermore, a sensitivity analysis revealed severe model dependency. While the optimised prompt achieved a 0.0% collision rate with the o3 model, it resulted in a 25.5% collision rate with Gemini-2.5-pro.

This research concludes that while base LLMs possess the emergent reasoning capabilities to function as high-level strategic agents, their lack of continuous perceptual flow limits their validity as direct operational controllers. The findings suggest that future implementations should adopt hierarchical architectures, leveraging LLMs for tactical reasoning while relying on physics-based controllers for dynamic execution. ... Read more

Automated vehicles (AVs) are expected to improve road safety, efficiency, and accessibility, yet their behaviour can at times appear overly cautious, rigid, or counter-intuitive, undermining trust and public acceptance. Existing approaches to address this problem, ranging from ethical decision-making models to behaviour imitation and interaction-based design, often lack a principled account of why certain behaviours should occur in specific contexts. This dissertation argues that these limitations stem from the absence of a unified framework that links human reasons to automated-vehicle decision-making in a transparent and evaluable manner.

To address this challenge, the thesis adopts the philosophical framework of Meaningful Human Control (MHC), which requires that automated systems both track relevant human reasons and allow responsibility for outcomes to be meaningfully traced to human agents. While MHC has been widely discussed at a conceptual level, its technical operationalisation in automated driving remains underdeveloped. This dissertation advances MHC by translating its normative principles into an integrated framework that connects ethical reasoning, engineering implementation, and empirical evaluation.

The dissertation first investigates which human reasons are relevant for automated-vehicle manoeuvre planning in ethically ambiguous, everyday traffic situations. Empirical findings from interviews with AV experts show that such reasons are inherently multi-layered, context-dependent, and often simultaneous, spanning normative, strategic, tactical, and operational considerations. Rather than functioning as fixed values or isolated cost terms, human reasons are shown to form context-sensitive relationships between underlying motivations and expected vehicle behaviour. These insights provide an empirically grounded basis for structuring and prioritising human reasons in automated-vehicle decision-making.

Building on this foundation, the dissertation develops a technical approach for embedding human reasons within automated-vehicle control architectures. Human reasons are translated into formal, machine-readable representations by drawing on insights from human-factors research and are integrated through a supervisory evaluation layer that operates alongside existing motion planning and control frameworks. This approach enables transparent trajectory evaluation and adaptive behavioural adjustment without requiring the design of new controllers, thereby demonstrating a practical pathway for operationalising MHC in real-time decision-making systems.

Finally, the dissertation examines whether meaningful human control can be empirically assessed in practice. Qualitative studies with users of partially automated driving systems reveal how the tracking and tracing conditions of MHC manifest dynamically in drivers’ experiences of safety, trust, responsibility, and intervention readiness. Complementary simulator experiments show that objective behavioural telemetry can capture aspects of tracking at the level of concrete interaction events, while tracing cannot be inferred from behaviour alone. Together, these findings demonstrate that meaningful human control is not merely a normative or post-hoc concept, but an empirically observable property of ongoing human–automation interaction when evaluated through a multi-layer framework combining subjective perception and objective data.

Overall, this dissertation advances the technical operationalisation of meaningful human control by systematically linking human reasons, automated-vehicle decision-making, and empirical evaluation. The proposed framework provides researchers, designers, and policymakers with concrete tools to assess and support reason-aligned automated-vehicle behaviour, contributing to the development of automated driving systems whose behaviour is more transparent, context-sensitive, and reasonable in everyday traffic situations.
... Read more

Annually, thousands of lives are lost to traffic accidents. To improve the safety of all traffic participants, the understanding and modelling of the limitations of human behaviour in traffic have continuously been researched. Currently, there is a lack of existing research on human predictions of other vehicles in traffic beyond binary decisions, such as whether the pedestrian will cross or whether another vehicle will accept the gap. This study conducted a human factors experiment with a novel response method where 30 participants viewed 168 unique scenarios for 5 seconds and then had to predict the intended direction the other vehicle would continue at the intersection. The direction predictions are a measure of how likely humans think the observed vehicle will go forward, left or right at the intersection. Analysis of the results showed that the heading angle and the relative position of the other vehicle had the greatest influence on the predicted direction and confidence of the response. Blinker use and deceleration had a lesser impact on prediction direction but significantly affected confidence. The lateral offset showed no statistical significance on the responses. The results highlight the limitations and inconsistencies in human predictions for other vehicles, particularly when the observed vehicle was positioned on the left or right side of an intersection, even when participants could focus solely on the other vehicle and no other distractions were present. Accounting for these inconsistencies when developing driving systems or testing autonomous vehicles can significantly enhance the safety and awareness of all traffic participants involved in an intersection. ... Read more

Human factor models has been an increasingly more popular topic in traffic models. The objective of these models vary, from simulating cooperative driving to understanding the behaviour of distracted drivers. Regardless of these diverse objectives, the reasons motivating these researches boil down to one single reason, safety. By better understanding human behaviour it should be possible to increase the safety of drivers on the road. One model which offers a systematic approach of studying human factors is the task-capability interface (TCI) model, it models the underlying human thought process and uses it as a proxy for other human factors. This has made the model quite successful in replicating various human factors, including distractions. Multiple papers have studied distractions using the TCI model as a tool but they all had their own specific approach to distractions. This leads to the identified gap in literature: how can distraction be systematically modelled in a TCI traffic model.\\
To fill this gap a distraction framework has been developed. This framework relies on the low-level characteristics of distractions and separates their lifecycle into three stages. These stages are the distraction trigger, intensity and effect. In order to verify if this framework is capable of systematically and accurately modelling distractions it was subjected to a validation test. To this end the distraction framework was incorporated into the Multi-scale model, which was found to be the most fitting TCI model, this resulted in the Distraction model. The new Distraction model was subsequently calibrated with a genetic algorithm to two different datasets with vastly different distraction, a continuous mental-visual distraction and a spontaneous auditory distraction. The results were compared to the results of specialized Distraction models.\\
The validation test results show that the Distraction model has shown limited improvements over the specialized baseline models and that most of the time its performance is equivalent. To be more specific the Distraction model is significantly better at estimating the headway of drivers compared to the baseline models when calibrating for single drivers. That said when it's used as a calibrated model it loses this edge and its performance is fully equivalent to the baseline models. With these results it can be concluded that the distraction framework functions as intended. Despite the limited amount of different distractions in the validation test it has shown that it is capable of systematically modelling distraction on a similar level as other specialized models. This also shows that the main benefits of the framework are its systematic approach and flexibility and not its performance capabilities. ... Read more

Strategy games provide a compelling testbed for developing human-like computer agents, with applications that extend beyond gaming into fields requiring adaptive and socially intelligent AI. In these games, players tend to enjoy and engage more deeply with AI opponents that not only provide a challenge but also behave in ways that resemble human thinking and decision-making. However, despite progress in developing such agents, there is still no standard approach for evaluating how human-like these opponents truly are—making it difficult to assess and improve their design. Here I show that strategy game opponents having more human-like game-level playstyles does not necessarily lead to them being more believable (perceived as human-like by human players).

By developing a turn-based strategy game and evaluating Hierarchical Reinforcement Learning (HRL) agents of varying complexity, I assessed both their behavioural similarity to human players and how believable they were perceived to be by human players. This research introduces a new approach for understanding player behaviour using behaviour vectors composed of three high-level metrics—Aggressiveness, Management, and Exploration—consistent with existing literature. These metrics are designed to be broadly applicable across strategy games, enabling consistent comparison between human and AI opponents, as well as across different games and agents. The findings demonstrate that while HRL agents can replicate human-like playstyles without using human training data, players judge human-likeness more on perceived intelligence and fairness. This suggests that creating truly human-like AI opponents requires not just replicating human game-level playstyles, but designing agents that align with players' expectations for intelligent and fair decision-making. ... Read more

Understanding how human drivers interact in dynamic traffic situations is a crucial step toward the safe and seamless integration of automated vehicles (AVs) into everyday traffic. A common setting for these interactions is the four way single-lane roundabout. Here, drivers must make quick decisions about who yields and who proceeds, based not just on traffic rules but also on subtle cues and shared expectations. These decisions rely heavily on gap acceptance, where each driver evaluates whether there is enough space and time to enter the roundabout safely. It often depends on mutual negotiation and split-second judgments, shaped by visual contact and behavioral feedback.
While earlier studies have explored driver gaze behavior in controlled environments, little is known about how gaze correlates with decision-making in continuous and mutual encounters, especially at roundabouts. This study fills that gap by studying human-human interactions during roundabout entry in a novel experimental setup. Using a coupled virtual reality driving simulator, two participants navigated a single-lane roundabout under varying approach speeds and distances. Eye-tracking was used to measure where and how long each driver fixated at the other vehicle. Control input data captured how drivers reacted in the seconds following these gaze events.
The results show that both entry distance and speed had a strong influence on who proceeded first. Drivers who started closer to the roundabout or moved faster were more likely to take priority. Drivers positioned closer to the conflict zone looked at the other vehicle for longer durations, indicating stronger visual engagement. Furthermore, drivers often responded with throttle or brake inputs shortly after looking at the other vehicle, especially when distance to the roundabout was small.
This study offers insight into how gaze behavior, positioning and control decisions shape mutual negotiation at roundabouts. These findings move beyond the idea of gap acceptance as a one-sided decision and highlight the importance of real-time interaction. ... Read more

With the increasing integration of Automated Vehicles (AVs) into our daily traffic, validating their performance poses a significant challenge. Virtual testing, where simulated AVs operate in a simulated environment, has become a widely adopted approach for efficient and cost-effective validation. However, the lack of realistic human behaviour models hinders the realism of these simulations, particularly in simulating reciprocal driver interactions. In this research, I introduce a discretionary lane change model based on the Communication-Enabled Interaction (CEI) framework, which simulates reciprocal driver interactions through implicit communication and belief modelling. These reciprocal interactions involve mutual behaviours, where individual drivers contribute through high-level decisions and low-level control actions. By employing the CEI framework, decision-making and control actions are integrated into a unified model. The proposed model is validated against naturalistic driving data in discretionary lane change scenarios to assess its validity. Results demonstrate that the model successfully reproduces qualitative and quantitative characteristics of human driving behaviour, reflecting both individual behaviours and the collective contributions of multiple drivers. Moreover, it reflects how varied tactical decisions yield distinct, human-like operational execution characteristics. Thereby improving the realism of interactive traffic simulations and posing a step towards improving virtual testing environments for AVs. ... Read more

Trajectory prediction is a key element of autonomous vehicle systems, enabling them to anticipate and react to the movements of other road users. Robustness testing through adversarial methods is essential for evaluating the reliability of these prediction models. However, current approaches tend to focus solely on manipulating model inputs, which can generate unrealistic scenarios and overlook critical vulnerabilities. This limitation may result in incomplete assessments of model performance in real-world conditions. The specific effects of more comprehensive adversarial attacks on trajectory prediction models have not been thoroughly investigated. In this work, we demonstrate that by perturbing both model inputs and anticipated future states, we can uncover previously undetected weaknesses and provide a more realistic evaluation of model robustness. Our novel approach incorporates dynamical constraints and preserves tactical behaviors, enabling more effective and realistic adversarial attacks. We introduce new performance measures to assess the realism and impact of these adversarial trajectories. Testing our method on a state-of-the-art prediction model reveals significant increases in prediction errors and collision rates under adversarial conditions. Qualitative analysis further shows that our attacks can expose critical weaknesses, such as the model’s inability to detect potential collisions in what appear to be safe predictions. These results underscore the need for more comprehensive adversarial testing to better evaluate and improve the reliability of trajectory prediction models for autonomous vehicles. To support further research in this area, we provide an open-source framework for studying adversarial robustness in trajectory prediction. This work advances adversarial testing techniques, contributing to the safety and reliability of autonomous driving systems. ... Read more

For trajectory prediction within autonomous vehicle planning and control, conditional variational autoencoders (CVAEs) have shown promise in accurate and diverse modeling of agent behaviors. Besides accuracy, explainability is also crucial for the safety and acceptance of learning-based autonomous systems, especially in autonomous driving. However, the latent distributions learned by CVAE models are often implicit and thus possess low explainability. To address this, we propose a semi-supervised generative modeling framework, \textbf{\textit{PrefCVAE}}, which utilizes partially and weakly labelled preference pairs to imbue the CVAE's latent representation with semantic meaning. This approach enables the system to estimate measurable attributes of the agents, and to generate manipulable predictions under the CVAE framework. Results show that incorporating our preference loss allows a CVAE-based model to make conditional predictions using the semantic factor of prediction average velocity. Our augmented framework also does not significantly degrade the baseline accuracy of prediction. Additionally, we show that the latent values learned with PrefCVAE better represent the semantic information contained in the data. Finally, we discuss the potential of this loss design to extend to other machine learning applications beyond trajectory prediction, as well as essential tricks for adaptation of human labeling. We hope that our empirical study offers the broader representation learning community a fresh perspective on inductive bias for disentangled and explainable latent representations in deep generative models. Specifically, we demonstrate that preference pair supervision, a simple and cost-effective approach, can effectively aid in learning semantic meanings for sampling-based generative models like the CVAE. ... Read more

Autonomy in traffic (e.g., autonomous vehicles) could potentially benefit mobility, safety, accessibility and sustainability. However, the realisation of these advancements is highly dependent on how effective these autonomous vehicles interact with vulnerable road users such as pedestrians. Before we can understand how pedestrians will interact with autonomous vehicles, it is essential to understand how pedestrians interact among themselves in interactive traffic scenarios. Previous studies have focused on describing these scenarios with probabilistic trajectory prediction methods such as TrajFlow. However, these approaches often fall short in capturing the nuances of mutual interactions. Simple interaction models have been proposed that can describe these interactions, but neglect the influence of another person's intentions. To address this issue, in existing work the Communication-Enabled-Interaction (CEI) framework was proposed that describes interactions by modelling communication and a belief of another person's intentions. The idea of using beliefs in interaction modelling is based on the concept that people have a general but uncertain idea about the plans of other people. These beliefs are one of the fundamental aspects of the CEI framework and must therefore contain valuable information about possible decisions. That is why this study investigates the use of the probabilistic trajectory prediction method TrajFlow for the belief construction of the CEI framework. TrajFlow is trained on the belief-based Forking Paths dataset, integrated into the CEI framework, and tested in four simulated pedestrian interaction scenarios. The analysis shows that the framework is able to simulate plausible interaction behaviour, dealing with conflicting goals and trajectories in multiple simulations. By doing so, this study takes a positive step towards modelling pedestrian interactions and contributes to the broader goal of realising the benefits linked to autonomy in traffic. ... Read more

This thesis explores enhancing track generalization in motorsport driver models through image-based feature sets, drawing inspiration from autonomous driving applications in urban settings. Traditional motorsport models often rely on numeric features, which excel on known tracks but face limita- tions when adapting to new, unseen environments. To address this, I introduce a CNN-based model that integrates bird’s- eye-view images with vehicle states and path-planning data, allowing a more holistic perception of track layouts and sur- roundings. Through open-loop evaluations on unseen tracks, the proposed model demonstrates superior generalization, achieving significantly lower RMSE compared to boundary point-based models, with improvements observed across steering, braking, and acceleration actions. Additionally, I apply novelty detection using Mahalanobis Distance to isolate Out-of-Distribution(OoD) scenarios, providing a precise measure of the generalization gap. This work establishes a baseline for image feature design in motorsport driver modeling, emphasizing the role of spatial and contextual information in achieving adaptable and high- performance autonomous racing agents. ... Read more

Automated driving technologies offer significant societal benefits but face challenges, particularly in interactions between automated and human-driven vehicles during lane changes and merging on highways. This thesis addresses this issue by focusing on joint driver efforts and proposes a new Communication-Enabled Interaction (CEI) model framework.

Human drivers communicate intent through vehicle kinematics during interactions, making joint decisions and exhibiting joint behaviors. However, current autonomous vehicle (AV) models often lack generalization to real-world behaviors and fail to capture dynamic interactions. AVs typically use models assuming human drivers only respond to AV behavior, leading to over-conservative and sometimes awkward interactions.

To enhance AV-human interactions, the thesis proposes a joint driver model that considers multi-level contributions of drivers. It critiques existing models, highlighting their limitations in capturing dynamic interactions. For instance, many models only consider single drivers and fail to address communication and continuous behavioral adaptation.

The CEI model framework explicitly accounts for driver communication and integrates deterministic future plans with probabilistic beliefs. This framework acknowledges that humans do not continuously optimize behavior but seek satisfactory solutions. The thesis presents a case study where the CEI model accurately describes merging scenarios, generating human-like gap-keeping behavior.

Further, the thesis explores naturalistic driving behaviors using the HighD dataset and develops visualization tools to validate driver models. It extracts and analyzes similar driving scenarios to understand variability in human responses, both operationally and tactically. Controlled experiments in simulators examine driver behaviors during merging conflicts, revealing insights into acceleration control and conflict resolution.

The empirical findings inspire improvements to the CEI model, incorporating intermittent piecewise-constant control observed in human drivers. This updated model accurately reproduces joint driver behaviors and outcomes from experimental scenarios, emphasizing the importance of individual contributions to joint safety margins.

In conclusion, the thesis contributes valuable insights into human lane-changing and merging interactions, proposing a robust model framework for AVs to understand and emulate human driver behaviors. While the study focuses on simplified scenarios, it lays the groundwork for extending the model to more complex real-world situations. The work represents a significant step toward enhancing autonomous vehicles' ability to interact safely and effectively with human drivers on the road. ... Read more

Understanding traffic participants’ behaviour is crucial for predicting their future trajectories, enabling autonomous vehicles to better assess the environment and consequently anticipate possible dangerous situations at an early stage. While the integration of cognitive processes and machine learning models has demonstrated promise in various domains, its application in trajectory forecasting of multiple traffic agents in large-scale autonomous driving datasets remains lacking. This work investigates the state-of-the-art trajectory forecasting model Trajectron++ which we enhance by incorporating a smoothing term in its attention module. This attention mechanism mimics human attention inspired by cognitive science research indicating limits to attention switching. We evaluate the performance of the resulting Smooth- Trajectron++ model and compare it to the original model on various benchmarks. Our results show improved performance on the large-scale nuScenes dataset, revealing the potential of incorporating insights from human cognition into trajectory prediction models. ... Read more

When a person makes a decision, it is automatically accompanied by a subjective probability judgement of the decision being correct, in other words, a (local) confidence judgement. Confidence judgements have, among other things, an
effect on justifications of future decisions and behaviour. A better understanding of the metacognitive processes responsible for these confidence judgements could improve behaviour models. To date, confidence judgements are mostly studied in a fundamental manner. Little to no research has been done into confidence in more dynamic tasks. Such applied research could render insights on whether fundamental principles also hold for real-life tasks. It could also have practical relevance for several applications. Driving is amongst the areas for which an improved understanding of the decision making and accompanied confidence judgements can be useful, for instance in order to improve driving assistance systems. However, current studies on driving behaviour are merely focused on decision making and do not take confidence into account.
In this study, we made a first attempt of connecting these two fields of research by investigating the confidence of drivers in left-turn gap acceptance decisions in a driver simulator experiment (N=17). The study showed that confidence can be related to the gap size with respect to the oncoming vehicle, described by the time-to-arrival and the distance gap. Confidence increases with the gap size for gap accepting decisions and decreases with the gap size for gap rejecting decisions. In addition, we concluded that confidence can be related to the driving behaviour, and that confidence is negatively related to the decision response time. Moreover, we found that confidence judgements can best be captured with the use of an extended dynamic drift diffusion decision model of which the drift rate of the evidence accumulator as well as the decision boundaries are functions of the time-to-arrival and distance gap. Furthermore, we demonstrated that allowing for post-decision evidence accumulation in the model increases its ability to describe confidence judgements in gap rejecting decisions. Overall, the study confirmed that principles known from fundamental confidence research can be used to describe confidence judgements in a dynamic and applied task. ... Read more

The rapid advancement in autonomous driving technology underscores the importance of studying the fragility of perception systems in autonomous vehicles, particularly due to their profound impact on public transportation safety. These systems are of paramount importance due to their direct impact on the lives of passengers and pedestrians. Additionally, their reliability can be easily compromised given the complexity and unpredictability of driving environments. However, current research and existing regulations often fail to adequately address the adversarial robustness of autonomous vehicle perception systems. This thesis delves into the adversarial robustness of camera-based perception systems of autonomous vehicles. Our research concentrates on developing and implementing evasion attacks that use black-box gradient estimation, as well as physical attacks in traffic sign detection and classification systems. Our findings indicate that even minor perturbations can impact the accuracy of these systems, leading to detection and classification errors. This finding highlights a critical vulnerability in the perception system's robustness against adversarial attacks. Moreover, the study extends to assess the transferability of adversarial examples across diverse perception models. Our results also expose significant gaps in the current regulatory frameworks of autonomous vehicles, necessitating the establishment of more rigorous and comprehensive safety standards. ... Read more

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. ... Read more

In order to design safe and effective interactions between autonomous vehicles (AVs) and human road users, it is essential to understand the mechanisms underlying human-human merging behavior. Driving simulator experiments can be used to study these mechanisms, but previous research has primarily focused on the behavior of individual drivers rather than the dynamics of interactions. In addition, current experimental scenarios and data analysis tools do not adequately capture interactive humanhuman merging behavior. To address these issues, I propose an experimental framework featuring a simplified highway-merging scenario that can facilitate human factors research on merging interactions. In a case study with fourteen participants, I used the framework in a coupled virtual reality driving simulator to show a relation between participants’ interactive behavior and fixation behavior. This work shows how to better understand human-human merging interactions, which is essential for developing AVs that can safely and successfully interact with other road users. ... Read more

Background: Merging on a highway is a complex driving task that requires a lot of interaction with other road users. During these tasks, a driver is required to evaluate gaps in space and time between the themselves and other road users and obstacles in order to arrive at the right moment to merge onto the highway. To improve safety and increase road efficiency, it is necessary to understand the decision process during merging decisions. A way of achieving this, is to understand what visual information humans use during this decision process. This study investigated the relation between gaze location and gap acceptance decisions during highway merges.
Methods: An experiment was performed in which 26 participants monitored an automated vehicle (AV) that was driving on a highway on-ramp. The participants were given the task to train the AV in whether or not to merge in front of an upcoming vehicle that was already driving on the highway. An eye tracker was used to measure gaze data, which was used to find the relation between gaze behaviour and decision outcomes and response times. A mixed-effects logistic model was used for a statistical analysis with decision outcomes as a dependent variable and different gap sizes as predictor variables. A mixed-effects linear model was used to find the relation between response times and dwell times and the different gap sizes and decision outcomes as predictor variables. For both the decision outcome and response time model, dwell time was later included to find the effect on the predictive validity.
Results: The results show that a larger time and distance gap to the upcoming vehicle relate to a higher merging probability. For larger time gaps to the on-ramp, the probability of merging was found to be smaller. It was also found that time gaps to the end of the on-ramp significantly relate to response times, with an increase of 55ms per 1s. Larger time gaps to the upcoming vehicle significantly relates to larger response times, with an increase of 64ms per 1s. No significant relation was found between response time and distance gaps to the upcoming vehicle. The response time was found to be 0.60s longer for rejected gap decisions. The time gap to the end of the on-ramp significantly relates to dwell time, with an increase of 0.56% per 1s. The distance gap to the upcoming vehicle significantly relates to dwell time, with an increase of 0.60% per 10m. The time gap to the upcoming vehicle significantly relates to dwell time, with an increase of 0.52% per 1s. The presented results show as well that a significant relation exists between gaze behaviour and decision outcomes and response times. When analysing decision outcomes and response times, the interaction between dwell time and gap sizes should be taken into account. This improved the predictive validity of the used regression models.
Conclusion: Several pieces of evidence suggest that gaze behaviour assist in understanding the human decision making process during merging. This study can serve as a basis for cognitive models that can investigate how the relation between gaze behaviour and gap sizes, decision outcomes and response times can help to understand and potentially predict gap acceptance decisions.
... Read more