Utilising raw psycho-physiological data and functional data analysis for estimating mental workload in human drivers

Abstract

Recent studies have focused on accurately estimating mental workload using machine learning algorithms and extracting features from physiological measures. However, feature extraction leads to the loss of valuable information and often results in binary classifications that lack specificity in the identification of optimum mental workload. This study investigates the feasibility of using raw physiological data (EEG, facial EMG, ECG, EDA, pupillometry) combined with Functional Data Analysis (FDA) to estimate the mental workload of human drivers. A driving scenario with five tasks was employed, and subjective ratings were collected. Results demonstrate that the FDA applied nine different combinations of raw physiological signals achieving a maximum 90% accuracy, outperforming extracted features by 73%. This study shows that the mental workload of human drivers can be accurately estimated without utilising burdensome feature extraction. The approach proposed in this study offers promise for mental workload assessment in real-world applications.

Practitioner summary

This study aimed to estimate the mental workload of human drivers using physiological signals and Functional Data Analysis (FDA). By comparing models using raw data and extracted features, the results show that the FDA with raw data achieved a high accuracy of 90%, outperforming the model with extracted features (73%).

RESEARCH HIGHLIGHTS

• The model utilising raw physiological signals achieved 90% accuracy, outperforming the model using extracted features with 73% accuracy.

• In the utilisation of extracted features for evaluation of mental workload of the human driver, Skin Conductance Level (SCL) greatly improved model performance in comparison with other features.

• The three optimal physiological signals for utilisation in real-time mental workload monitoring of the human driver were ECG, EDA, and eye tracking data.

• FDA offers interpretable results, providing insights into the progression of mental workload and individualised prediction outcomes.

Keywords:

• Bio-signals
• Psycho-physiological
• physiological
• mental workload
• functional data analysis
• human driver
• subjective ratings
• estimation model
• model comparison

Disclosure statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Additional information

Funding

The research was funded by the MSIT (Ministry of Science, ICT), Republic of Korea, under the High-Potential Individuals Global Training Program (RS-2022-00154879) supervised by the IITP (Institute for Information &Communications Technology Planning & Evaluation); and University of Windsor through the Innovating Sustainability Grant program (822193). Kyongwon Kim was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (RS-2023-00219212). The authors thank Dr. Umair Durrani for development of driving simulator scenarios in this study.