Understanding cyclists' perception of driverless vehicles through eye-tracking and interviews

Overview

• Researchers investigated how cyclists interact with automated vehicles (AVs).
• They looked at whether cyclists notice if a vehicle is driverless, how well they can detect the presence or absence of a driver, and how they carry out such detection tasks.
• 37 participants cycled a designated route and encountered an AV multiple times in two sessions.
• In the first session, they cycled without instructions, while in the second session, they were asked to verbally report whether they detected a driver.
• The researchers used eye-tracking to record the participants' gaze behavior and conducted post-session interviews.

Plain English Explanation

As automated vehicles (AVs) become more common, researchers wanted to understand how cyclists interact with them. They set up an experiment where cyclists rode a route and encountered an AV multiple times.

In the first part of the experiment, the cyclists just rode the route without any instructions. The researchers found that 30% of the cyclists spontaneously noticed that the vehicle didn't have a driver. In the second part, the cyclists were asked to report whether they saw a driver or not. When given this instruction, they were able to correctly detect the presence or absence of a driver 93% of the time.

The researchers also used eye-tracking technology to see where the cyclists were looking when they encountered the AV. They found that the cyclists looked at the vehicle more often and for longer periods of time when they were instructed to detect the driver, compared to when they were just riding normally.

The cyclists also exhibited a specific pattern in how they looked at the vehicle. When it was far away, they tended to look in front of it, but when it got closer, they focused more on the windshield area. The cyclists reported feeling curious and safe around the AV, and they said they would like to receive more information about the vehicle's driving state.

Overall, this study shows that cyclists can detect when a vehicle is driverless, and this awareness may influence their perceptions of safety. Further research is needed to see how these findings play out in real-world traffic conditions.

Technical Explanation

The researchers used a Wizard-of-Oz method to investigate how cyclists interact with AVs. In this approach, the researchers used a human driver hidden in the vehicle to simulate an AV, rather than using a fully autonomous vehicle.

A total of 37 participants cycled a designated route and encountered the AV multiple times in two experimental sessions. In the first session, the participants cycled the route without any instructions. The researchers found that 30% of the cyclists spontaneously mentioned the absence of a driver in the post-session interviews.

In the second session, the participants were instructed to verbally report whether they detected the presence or absence of a driver in the AV. Under these instructions, the cyclists were able to correctly identify the presence or absence of a driver with 93% accuracy.

The researchers also used eye-tracking technology to record the participants' gaze behavior. The data showed that the cyclists looked more frequently and for longer durations at the vehicle in the second session, when they were explicitly instructed to detect the driver, compared to the first session.

Furthermore, the eye-tracking data revealed a specific pattern in how the cyclists looked at the vehicle. When the AV was far away, the cyclists tended to look in front of the vehicle, but when it was closer, they focused more on the windshield region.

The post-session interviews also provided insights into the cyclists' perceptions and experiences. The participants reported feeling curious, safe, and expressed a desire to receive more information about the AV's driving state.

Critical Analysis

The study provides valuable insights into how cyclists interact with AVs, but it also has some limitations that should be considered.

First, the use of a Wizard-of-Oz method, while necessary for this type of research, may not fully capture the complexities of real-world interactions between cyclists and fully autonomous vehicles. The presence of a hidden human driver could influence the cyclists' perceptions and behaviors in ways that may not translate to actual AV encounters.

Additionally, the study was conducted in a controlled, designated route, which may not reflect the chaotic and unpredictable nature of real-world traffic conditions. Further research is needed to explore these findings in more realistic settings, where cyclists may face a wider range of scenarios and have to make split-second decisions.

The study also focused on a relatively small sample size of 37 participants. While the results provide valuable insights, larger-scale studies would be helpful to validate the findings and explore potential individual differences or cultural factors that may affect cyclist-AV interactions.

Finally, the research did not investigate the long-term implications of cyclists' ability to detect the absence of a driver. It would be interesting to explore how this awareness might shape their trust, acceptance, and overall interactions with AVs over time, as the technology becomes more widespread.

Conclusion

This study suggests that cyclists can readily detect when a vehicle is driverless, and this awareness may influence their perceptions of safety and their interactions with AVs. The findings highlight the importance of considering the needs and behaviors of cyclists in the development and deployment of autonomous vehicle technologies.

As AVs become increasingly prevalent, it will be crucial to further investigate the dynamic interplay between cyclists and these vehicles, particularly in real-world traffic conditions. By understanding how cyclists interact with and perceive AVs, researchers and policymakers can work towards creating safer and more inclusive transportation systems that accommodate the needs of all road users.