A Dataset on Takeover during Distracted L2 Automated Driving

Area of Science:

• Human-Computer Interaction
• Automotive Engineering
• Cognitive Psychology

Background:

• Automated driving systems (ADS) allow drivers to engage in non-driving tasks, raising concerns about driver distraction.
• Driver distraction during ADS operation necessitates research into driver performance during the critical fallback to driving (takeover) process.
• Existing publicly available datasets are insufficient for advancing the development of safe ADS, particularly concerning takeover performance.

Purpose of the Study:

• To introduce TD2D, a comprehensive dataset designed to support research in automated driving systems.
• To provide a rich dataset encompassing takeover performance, workload, physiological, and ocular metrics.
• To facilitate advancements in the safety and design of automated driving technologies.

Main Methods:

• Collected data from 50 drivers (balanced gender, diverse age groups) in an L2 automated driving simulator.
• Recorded 500 takeover cases across 10 distinct secondary task conditions (visual, auditory, and none).
• Acquired takeover performance, workload, physiological, and ocular data for each case.

Main Results:

• The TD2D dataset includes detailed metrics on driver responses during takeover scenarios.
• Data covers a range of driver engagement levels due to varied secondary task complexities.
• The dataset captures critical physiological and ocular indicators related to driver attention and workload.

Conclusions:

• The TD2D dataset represents a significant contribution to the field of automated driving research.
• This dataset will enable more robust analysis of driver behavior and performance during takeover.
• TD2D is expected to accelerate the development of safer and more reliable automated driving systems.