Secondary Tasks and Autonomous Vehicles

Grade Level at Time of Presentation

Senior

Major

Psychology

Minor

Business

2nd Grade Level at Time of Presentation

Senior

2nd Student Major

Psychology

3rd Grade Level at Time of Presentation

Junior

3rd Student Major

Psychology

Institution

Morehead State University

KY House District #

96

KY Senate District #

18

Department

Psychology

Abstract

Self-driving vehicles are now rising in popularity. It is essential that we understand and evaluate the risks of this new technology. Autonomous vehicles depend on the human operator to monitor and respond appropriately when the automation fails. We used a tracking task to simulate a vehicle driving down a road. In autonomous mode, the cursor in the tracking task, mimicking an autonomous vehicle, stayed in a region of a computer monitor. Once the autonomous mode failed the cursor strayed out of that region. The participant’s task was to monitor the cursor’s position, and once the automation failed, to bring the cursor back under control. In autonomous mode, the driver of a vehicle is free to do other, secondary, tasks. For our simulation, participants solved arithmetic problems, presented either auditorily and visually, as the secondary task. Fourteen female participants were recruited to participate in the study. All participants were treated in accordance to IRB protocol. The dependent variable was the deviation from the target area, measured in RMSE. The independent variables were the two modes of secondary task presentation, audio and visual. Presentation of the secondary task was a between participant variable. We also divided the total time of the non-autonomous tracking task into 8 segments of 15 seconds. For RMSE the visual task was more disruptive than the auditory task. There was no interaction between the time segments and the type of the secondary task. The primary scientific concern was the length of time to regain control of the vehicle and whether that time differed as a function of the mode for the secondary task. The time to regain control improved over the 2-minute segment of manual control, and the auditory secondary task resulted in the smallest RMSE. There were no interactions.

This document is currently not available here.

Share

COinS
 

Secondary Tasks and Autonomous Vehicles

Self-driving vehicles are now rising in popularity. It is essential that we understand and evaluate the risks of this new technology. Autonomous vehicles depend on the human operator to monitor and respond appropriately when the automation fails. We used a tracking task to simulate a vehicle driving down a road. In autonomous mode, the cursor in the tracking task, mimicking an autonomous vehicle, stayed in a region of a computer monitor. Once the autonomous mode failed the cursor strayed out of that region. The participant’s task was to monitor the cursor’s position, and once the automation failed, to bring the cursor back under control. In autonomous mode, the driver of a vehicle is free to do other, secondary, tasks. For our simulation, participants solved arithmetic problems, presented either auditorily and visually, as the secondary task. Fourteen female participants were recruited to participate in the study. All participants were treated in accordance to IRB protocol. The dependent variable was the deviation from the target area, measured in RMSE. The independent variables were the two modes of secondary task presentation, audio and visual. Presentation of the secondary task was a between participant variable. We also divided the total time of the non-autonomous tracking task into 8 segments of 15 seconds. For RMSE the visual task was more disruptive than the auditory task. There was no interaction between the time segments and the type of the secondary task. The primary scientific concern was the length of time to regain control of the vehicle and whether that time differed as a function of the mode for the secondary task. The time to regain control improved over the 2-minute segment of manual control, and the auditory secondary task resulted in the smallest RMSE. There were no interactions.