Interaction Effects of Pedestrian Behavior, Smartphone Distraction and External Communication of Automated Vehicles on Crossing and Gaze Behavior

Title of the Paper

Interaction Effects of Pedestrian Behavior, Smartphone Distraction, and External Communication of Automated Vehicles on Crossing and Gaze Behavior

Paper Information

• Subject Area: Human-Computer Interaction Research in Transportation and Autonomous Driving Technology
• Keywords: Automated Vehicles, External Human-Machine Interface (eHMI), Virtual Reality, Pedestrian Groups, Smartphone Distraction, Unsignalized Crosswalks, Eye Tracking

Research Background and Problem

• Identified Problems or Challenges: As automated vehicles (AVs) gradually replace traditional vehicles, direct communication between drivers and pedestrians is disappearing. In complex traffic scenarios (e.g., multiple pedestrians crossing simultaneously or pedestrians distracted by smartphone use), existing research has not sufficiently explored the interaction effects of these factors on pedestrian behavior.
• Importance of the Problem: Automated vehicles need reliable methods to convey their intentions to pedestrians to ensure safe crossing behavior. The successful application of these technologies is crucial for reducing traffic accidents and addressing uncertainties in complex traffic scenarios.
• Motivation and Related Work: Although previous studies have evaluated the impact of external human-machine interfaces (eHMI) of automated vehicles on pedestrian safety, these studies primarily focus on simple scenarios involving a single pedestrian and a single vehicle, neglecting the complexities of distracted pedestrians and multi-pedestrian scenarios.

Solution

• Proposed Solution:

  • Design and explore the role of external communication methods (eHMI) of automated vehicles in complex scenarios through virtual reality (VR) experiments.
  • Investigate the interaction effects of smartphone-distracted pedestrians, pedestrian groups, and automated vehicles using eHMI.

• Innovative Contributions:

  • Incorporating pedestrian distraction and group behavior into the study, and for the first time, exploring the interaction effects of these factors with the external communication effectiveness of automated vehicles.
  • Providing research scenarios that are closer to real-world conditions rather than relying solely on controlled laboratory environments.

• Implementation Steps:

  1. Design VR experiments, including simulated streets and visual eHMI for automated vehicles (e.g., LED light strips).
  2. Assign participants smartphone distraction tasks and introduce virtual pedestrian groups.
  3. Collect objective data such as crossing time, crossing duration, and gaze behavior, while also evaluating subjective factors like pedestrians' perceived safety and cognitive load.

Research Findings

• Specific Findings:

  • Objective Performance: The positive effects of eHMI (e.g., faster crossing) significantly decrease when pedestrians are distracted. The presence of pedestrian groups has a positive impact on distracted pedestrians, reducing their cognitive load and increasing their sense of safety.
  • Subjective Evaluation: Distracted pedestrians perceive the effectiveness of eHMI to be weaker. However, in the presence of pedestrian groups, distracted pedestrians report higher subjective safety and lower perceived situational criticality.
  • Gaze Behavior: Distracted pedestrians spend less time looking at traffic, but when eHMI is active, they focus more attention on traffic. In the presence of pedestrian groups, they pay less attention to stationary vehicles.

• Comparison with Existing Solutions:

  • Previous studies found that eHMI has positive effects on pedestrians only when they are not distracted.
  • This study reveals distinct behavioral patterns of distracted pedestrians in complex traffic scenarios, addressing gaps in existing research.

• Experimental Results and Limitations:

  • The results demonstrate the impact of eHMI and pedestrian groups on the crossing behavior of distracted pedestrians.
  • Limitations include the potential inability of VR scenarios to fully replicate real-world environments, and the participant sample being predominantly young, which limits generalizability to other age groups.

• Future Directions:

  • Investigate whether eHMI designs based on smartphones or ground displays can improve the safety of distracted pedestrians.
  • Expand scenarios to include more pedestrians, multiple vehicles, and other types of road users (e.g., cyclists and motorcyclists).
  • Conduct experiments with elderly or special groups to explore their behavior patterns in complex traffic scenarios.

Conclusion

This study integrates smartphone distraction, pedestrian group behavior, and the external communication methods of automated vehicles to explore their interaction effects. The results indicate that eHMI is less effective for distracted pedestrians in complex scenarios, while pedestrian groups provide psychological and behavioral benefits to distracted individuals. This research provides a critical foundation for designing more effective communication technologies for automated vehicles and for understanding complex traffic scenarios.