Programs

Invited presentation 1 (15min+5min)

Continual Driver Behavior Learning In Interactive Urban Scenarios 

Abstract:

Developing autonomous vehicles (AVs) helps improve the road safety and traffic efficiency of intelligent transportation systems (ITS). Accurately predicting the trajectories of traffic participants is essential to the decision-making and motion planning of AVs in interactive scenarios. Recently, learning-based trajectory predictors have shown state-of-the-art performance in highway or urban areas. However, most existing learning-based models trained with fixed datasets may perform poorly in continuously changing scenarios. Specifically, they may not perform well in learned scenarios after learning the new one. This phenomenon is called "catastrophic forgetting". Few studies investigate trajectory predictions in continuous scenarios, where catastrophic forgetting may happen. To handle this problem, first, a novel continual learning (CL) approach for vehicle trajectory prediction is proposed in this paper. Then, inspired by brain science, a dynamic memory mechanism is developed by utilizing the measurement of traffic divergence between scenarios, which balances the performance and training efficiency of the proposed CL approach. Finally, datasets collected from different locations are used to design continual training and testing methods in experiments. Experimental results show that the proposed approach achieves consistently high prediction accuracy in continuous scenarios without re-training, which mitigates catastrophic forgetting compared to non-CL approaches.

Abstract:

We introduce our project to promote wellbeing in future hybrid mobility society where humans and autonomous agents seamlessly interact in a shared space.  We believe a sense of equality and happiness is vital for smooth interaction in such a society. Society’s collective well-being reflects the quality of life of every single individual including physical and emotional fulfillment, along with self-realization in a harmonious setting. To enhance users’ happiness and satisfaction, we envision future technologies that promote humans to be pro-social through interaction with automated agents. We have developed mobility simulators to analyze user interactions in hybrid mobility environment and conducted several user studies to evaluate user responses. In this talk, we present our methodology and initials findings demonstrating how users’ wellbeing can be affected during pro-social interactions with autonomous agents.

Invited presentation 3 (15min+5min)

Is Silent eHMI Enough? 

A Passenger-Centric Study on Effective eHMI for Autonomous Personal Mobility Vehicles in the Field

Abstract:

Autonomous personal mobility vehicle (APMV) is a miniaturized autonomous vehicle that is used in pedestrian-rich environments. In addition, the open bodywork design of APMVs exposes passengers to the communication between the eHMI deployed on APMVs and pedestrians. Therefore, to ensure an optimal passenger experience, eHMI designs for APMVs must consider the potential impact of APMV-pedestrian communications on passengers' psychological feelings. This study discussed three external human-machine interface (eHMI) designs, i. e., eHMI with text message (eHMI-T), eHMI with neutral voice (eHMI-NV) and eHMI with affective voice (eHMI-AV), from the perspective of APMV passengers in the communication between APMV and pedestrians. In the riding experiment, we found that eHMI-T may be less suitable for APMVs. This conclusion was drawn based on passengers' feedback, as they expressed feeling awkward during the "silent period" because the eHMI-T conveyed information only to pedestrians but not to passengers. Additionally, the affective voice cues on eHMI improved overall user experience of passengers, leading to higher ratings for both pragmatic and hedonic quality. The study also highlights the necessity of considering passengers' personalities when designing eHMI for APMVs to enhance their experience.

Workshop paper 1 (15min+5min)

You Cooperate, I Reciprocate: 

Well-Being and Trust in Automated Vehicles

Mengyao Li 1, Shashank Mehrotra 2, Kumar Akash 2, Teruhisa Misu 2, John Lee 1

1. University of Wisconsin-Madison, Madison, Wisconsin, USA

2. Honda Research Institute USA, Inc., San Jose, California, USA

Abstract:

Cooperative automated vehicles (AVs) bring the potential for better safety, efficiency, and energy-savings on the individual and system level. Yet, these benefits can only be achieved if people cooperate. In this study, we explored the effects of cooperative and reciprocal AVs on people’s well-being, trust, and cooperation. We conducted a mixed-design study (n = 304), where participants experienced 4 types of social interactions as between-subject conditions: 1) all altruistic interaction; 2) all selfish interaction; 3) altruistic other vehicle, selfish ego AV; 4) selfish other vehicle, altruistic ego AV. We found people’s well-being was highest when other vehicle is selfish and ego AV is altruistic; whereas people’s trust was the highest when people experienced all altruistic interactions. Results suggested a design balance when evaluating people’s attitudes towards AVs: altruistic AVs can promote people’s trust, whereas when evaluating people’s well-being, the presence of selfish AVs may be beneficial. Future studies could model the balance between the user optimal and system optimal.

Workshop paper 2 (15min+5min)

Understanding User Acceptance of Shared Autonomous Vehicles for Potential Use in Public Transport: 

Insights from Focus Groups in Sweden

Sigma Dolins 1,2, MariAnne Karlsson 1, Helena Strömberg 1

1. Department of Design & Human Factors at Chalmers University of Technology, Sweden

2. RISE Research Institutes of Sweden, Sweden

Abstract:

This paper investigates the acceptance of shared, autonomous vehicles (AVs) with Swedish participants who have used or live near AV pilots, with a specific focus on shared mobility and public transport contexts. The study aims to gain insights into individuals' attitudes, acceptance levels, and willingness to use AVs in shared mobility scenarios, particularly in the context of public transportation. The research methodology employed focus groups to gather qualitative insights from participants. The findings provide valuable insights into user perceptions and preferences, offering guidance for promoting shared AVs and integrating them into public transport systems in Sweden.

Invited presentation 4 (15min+5min)

Computational Models of Motion Sickness and Their Application to Comfort in Automated Vehicles

Abstract:

As long as automated driving vehicles are developed for transporting people, it is essential that they provide a comfortable experience for the passengers. The NAIST Human Robotics Lab has been conducting research from various perspectives. My talk focuses on a computational model for quantifying motion sickness and their applications to countermeasures to reduce discomfort for passengers of automated vehicles. In addition, if time is available, I might also talk about other topics related to human factors of automated vehicles.

Invited presentation 5 (15min+5min)

Motion comfort, perceived safety and trust in automated driving

Abstract:

Motion comfort, the perception of safety and trust in automation are pivotal factors in the acceptance of higher automation levels. This presentation will present human models predicting these factors including the immediate perception of motion comfort, the accumulation of motion sickness in time, perceived risk in critical events, and the development of trust. Examples will be presented how vehicle motion control can enhance motion comfort, and how interfaces showing the user what the car perceives and explaining vehicle manoeuvres can enhance perceived safety and trust in automation.

Invited presentation 6 (15min+5min)

Earth-Fixed Books Reduce Motion Sickness When Reading with a Head-Mounted Display

Abstract:

There are concerns that viewing two-dimensional (2D) content such as web pages on a head-mounted display (HMD) in the car may aggravate motion sickness. This is because when 2D content is fixed to a head-fixed coordinate system, the appearance of the content does not change even when the body moves; therefore, it is impossible to visually perceive the movement of one’s body, resulting in a sensory conflict between the visual and vestibular senses. A method for reducing motion sickness when displaying 3D content on an HMD has been investigated; however, when displaying 2D content, no such method has been investigated. Therefore, this study aims to verify to the possibility of reducing motion sickness from the change of appearance caused by fixing 2D content to the earth-fixed coordinate system when viewing it with an HMD in a moving environment. Participants sat on a seat that was mounted on a vibrating device and moved in the pitch direction while reading a book on the HMD. Consequently, the severity of motion sickness was significantly lower when the book was fixed to the earth-fixed coordinate system than when fixed to the head-fixed coordinate system. This result suggests that by fixing the content to the earth-fixed coordinate system, motion sickness can be reduced because the movement of one’s body can be perceived through changes in the appearance of the content, and the sensory conflict between visual and vestibular sensations can be resolved.

Invited presentation 7 (15min+5min)

Improving Pedestrian Priority

via Grouping and Virtual Lanes

Abstract:

The concept of shared space design has been increasingly adopted in urban streets as a means to facilitate barrier-free movement and foster the integration of various traffic participants, including pedestrians, cyclists, and motor vehicles, within a unified road space. Despite the advantages of low-speed environments, sharing the road with motor vehicles can instill feelings of unease and vulnerability among pedestrians. However, research has shown that walking in groups can enhance their confidence while also influencing the yielding behavior of drivers. In response to this challenge, we propose an innovative approach to enhance pedestrian crossing in shared spaces by leveraging grouping techniques and projecting virtual lanes. We will introduce the fundamental components of the crowd steering system, comprising generating virtual lanes and its different visualization, as viable means to facilitate pedestrian grouping and improve their crossing experience. Additionally, we delve into a comprehensive discussion of the enablers and current gaps in the existing approach, shedding light on the potential challenges and opportunities for refinement.

Workshop paper 3 (15min+5min)

Integrating Generative Artificial Intelligence in Intelligent Vehicle Systems: A Research Agenda

Lukas Stappen, Jeremy Dillmann, Serena Striegel, 

Hans Joerg Voegel, Nicolas Flores-Herr, Björn Schuller

BMW Group Research and Technology, Germany

Abstract:

This paper aims to serve as a comprehensive guide for researchers and practitioners, offering insights into the current state, potential applications, and future research directions for generative artificial intelligence and foundation models within the context of intelligent vehicles. As the automotive industry progressively integrates AI, generative artificial intelligence technologies hold the potential to revolutionize user interactions, delivering more immersive, intuitive, and personalized in-car experiences. We provide an overview of current applications of generative artificial intelligence in the automotive domain, emphasizing speech, audio, vision, and multimodal interactions. We subsequently outline critical future research areas, including domain adaptability, alignment, multimodal integration and others, as well as, address the challenges and risks associated with ethics. By fostering collaboration and addressing these research areas, generative artificial intelligence can unlock its full potential, transforming the driving experience and shaping the future of intelligent vehicles.

Workshop paper 4 (15min+5min)

Modeling gap acceptance in overtaking: 

a cognitive process approach

Samir H.A. Mohammad 1, Haneen Farah 2, Arkady Zgonnikov 1

1.Departmentof Cognitive Robotics, Faculty of Mechanical, Maritime and Materials Engineering, Delft University of Technology, the Netherlands

2. Department of Transport and Planning, Faculty of Civil Engineering and Geosciences, Delft University of Technology, the Netherlands

Abstract:

Driving automation holds significant potential for enhancing traffic safety. However, effectively handling interactions with human drivers in mixed traffic remains a challenging task. Several models exist that attempt to capture human behavior in traffic interactions, often focusing on gap acceptance. However, it is not clear how models of an individual driver’s gap acceptance can be translated to dynamic human-AV interactions in the context of high-speed scenarios like overtaking. In this study, we address this issue by employing a cognitive process approach to describe the dynamic interactions by the oncoming vehicle during overtaking maneuvers. Our findings reveal that by incorporating an initial decision-making bias dependent on the initial velocity into existing drift-diffusion models, we can accurately describe the qualitative patterns of overtaking gap acceptance observed previously. Our results demonstrate the potential of the cognitive process approach in modeling human overtaking behavior when the oncoming vehicle is an AV. To this end, this study contributes to the development of effective strategies for ensuring safe and efficient overtaking interactions between human drivers and AVs.