Think Aloud Protocol Applied in Naturalistic Driving for Driving Rules Generation

Area of Science:

• Human-Computer Interaction
• Autonomous Driving Systems
• Cognitive Psychology

Background:

• Modeling driver behavior is crucial for advancing autonomous driving technology.
• Existing methods for driver behavior modeling have limitations in complex scenarios.
• Naturalistic driving studies provide realistic data for understanding human driving.

Purpose of the Study:

• To present the Think Aloud Protocol as a methodology for modeling driver behavior.
• To demonstrate the application and validity of the Think Aloud Protocol in naturalistic driving.
• To derive driving rules from driver verbalizations in complex scenarios, specifically roundabouts.

Main Methods:

• The Think Aloud Protocol involves drivers verbalizing their thoughts while driving.
• Data collection includes recording audio and video of drivers in naturalistic settings.
• Analysis involves processing recordings to generate coded transcript files and derive driving rules.

Main Results:

• The Think Aloud Protocol is applicable to naturalistic driving research.
• The methodology facilitates the derivation of specific driving rules, as demonstrated in roundabout scenarios.
• The study identified potential limitations of the Think Aloud Protocol in this context.

Conclusions:

• The Think Aloud Protocol offers a viable and relatively easy approach to derive driving rules from naturalistic driving data.
• This methodology contributes to a deeper understanding of driver behavior, essential for safe autonomous systems.
• Further research can refine the protocol to address identified limitations and enhance its application.