The advancements in driverless car technology have paved the way for vehicles with auto-pilot capabilities. While these cars offer a more convenient driving experience, there are concerns regarding the readiness of drivers to respond to warning signals when engaged in other activities. A recent study by researchers from the University College London (UCL) explores the possibility of using eye movements to determine drivers’ attention levels and their ability to respond to takeover signals.
The research published in Cognitive Research: Principles and Implications aims to shed light on how engrossed drivers are in on-screen activities while using auto-pilot mode. The study suggests that through eye movement analysis, it is possible to assess whether drivers are sufficiently attentive to respond to real-world signals, such as takeover requests from the car.
To investigate the relationship between attention levels and eye movements, the researchers conducted experiments involving 42 participants. The participants were asked to complete search tasks on a computer screen while their gaze patterns were monitored. The tasks varied in difficulty, ranging from identifying simple shapes to recognizing specific arrangements and colors.
The researchers found that participants took longer to stop watching the screen and respond to a tone when the task demanded more attention. By analyzing the eye movement patterns, a correlation was observed between longer fixations and shorter distances of eye travel. These eye movement patterns indicated higher attention demands during the task.
Building on these findings, the researchers trained a machine learning model using the collected data to predict participants’ engagement levels in easy or demanding tasks based solely on their eye movement patterns. The findings revealed that the model could accurately predict the level of task engagement from eye movements alone.
The development of autonomous vehicles holds the promise of enhancing driving experiences and enabling drivers to engage in other non-driving tasks during their commute. However, a key concern lies in the ability of drivers to swiftly return to full driving control when receiving a takeover signal.
The study’s senior author, Professor Nilli Lavie, emphasizes the importance of understanding drivers’ attention levels and readiness to respond to warning signals. The research demonstrates that monitoring gaze patterns can effectively assess drivers’ attention levels and their ability to respond promptly.
While the results of this study provide valuable insights, the researchers acknowledge the need for larger datasets to further refine and improve the machine learning model. Expanding the dataset will help enhance the accuracy of predictions and ensure the reliability of the findings.
The study conducted by UCL researchers highlights the potential of eye movement analysis in evaluating driver attention levels while using auto-pilot mode. By understanding how drivers’ attention is affected by on-screen activities, future advancements in driverless car technology can prioritize safety without compromising the convenience and productivity offered by auto-pilot capabilities. Further research and development are necessary to refine the methods and incorporate them into the design of autonomous vehicles, ensuring the seamless transition between non-driving tasks and full driving control.