Driver's Cognitive Function Estimation Using Daily Driving Data.

Area of Science:

• Human-computer interaction
• Cognitive science
• Transportation engineering

Background:

• Driving assistance systems can enhance safety by adapting to driver characteristics.
• Estimating drivers' cognitive function from daily driving data can support older drivers without burden.
• Cognitive function, measured by Trail Making Test (A) and (B), is crucial for driving ability.

Purpose of the Study:

• To develop a method for estimating drivers' cognitive function using everyday driving data.
• To model the relationship between driving behaviors and cognitive function scores.
• To investigate the utility of specific driving scenarios for cognitive function assessment.

Main Methods:

• Utilized driving behavioral features extracted from diverse driving routes.
• Focused on particular driving scenarios where cognitive differences are more apparent.
• Modeled Trail Making Test (A) and (B) scores as indicators of cognitive function.

Main Results:

• Achieved Spearman rank correlation coefficients of r = 0.34 for Trail Making Test (A) and r = 0.48 for Trail Making Test (B).
• Demonstrated the effectiveness of incorporating driving scenario information into the model.
• Facilitated the analysis of relationships between specific driving behaviors (e.g., steering angle velocity) and cognitive function in scenarios like intersections.

Conclusions:

• Daily driving data can be used to estimate drivers' cognitive function.
• The proposed method, focusing on specific scenarios, improves the accuracy of cognitive function estimation.
• This approach offers a non-intrusive way to assess cognitive status for adaptive driving systems.