A multimodal driver monitoring benchmark dataset for driver modeling in assisted driving automation

Khazar Dargahi Nobari ¹, Torsten Bertram ²

⁰TU Dortmund University, Institute of Control Theory and Systems Engineering, Otto-Hahn-Str. 8, 44227, Dortmund, Germany. khazar.dargahinobari@tu-dortmund.de.

Scientific Data +

|

March 30, 2024

View abstract on PubMed

Summary

The manD 1.0 dataset offers synchronized multimodal data from 50 participants in a driving simulator. This benchmark supports driver monitoring research for automated driving systems.

Area Of Science

Human-Computer Interaction
Automated Driving Systems
Behavioral Science

Background

Driver monitoring is crucial for interpreting, modeling, and predicting driver behavior in automated driving.
Existing datasets may lack the multimodal synchronization required for comprehensive analysis of driver states.

Purpose Of The Study

Introduce manD 1.0, a novel multimodal dataset for benchmarking driver monitoring in automated driving.
Provide a standardized resource for research into human dimension in automated driving (manD).

Main Methods

Collected synchronized data from 50 gender-balanced participants (ages 21-65) in a static driving simulator.
Simulated five distinct driving scenarios with varying automation levels (SAE L0-L3), traffic, and weather conditions.
Captured multimodal driver data including physiology, body movements, activities, gaze, and facial information.

Main Results

The manD 1.0 dataset contains synchronized environmental, vehicle, and comprehensive driver state data.
Data reflects diverse mental and physical states across various driving events and conditions.
The dataset is suitable for developing and validating driver monitoring applications.

Conclusions

manD 1.0 serves as a valuable benchmark for driver monitoring research in automated driving.
The dataset facilitates data-driven modeling, prediction of driver reactions, and interaction strategy design.
Enables further research into driver states, including motion sickness, within automated driving contexts.