Autonomous Vehicle and Cybernetics

Research

Inferring driver workload with individualization using physiological data

The ability to measure and detect driver’s workload has been an important research topic in automotive research domain, and started to attract greater interest in recent years in relation to the emerging vehicle technology such as autonomous driving. Our study incorporates the individual difference to characterize the workload response on personal basis. Extensive pre-processing of electroencephalogram (EEG) and electrocardiogram (ECG) signals were carried out to generate features for personalization. The results showed that human workload response is far from homogeneity, nor the driving environment is a sole determinant.

SELECTED REFERENCES

Yuna Noh, "Modeling Individual Differences in Driver Workload Estimation using Physiological Data", Ph.D Thesis, 2018

Yuna Noh, Yoonjin Yoon, "Personalized stress assessment during real-world electric vehicle driving based on fuzzy c-means clustering using physiological and operational data", under review

Characterizing driver stress using physiological and operational data

From 53 km Electric vehicle driving experiment of 40 subjects, physiological data including electroencephalogram (EEG) and eye-gazing were obtained along with vehicle operational data such as state of charge, altitude, and speed. The dataset was rich in information, but individual difference and nonlinear patterns made it extremely difficult to draw meaningful insights. As a solution, an information-theoretic framework is proposed to evaluate mutual information between physiological and operational data as well as the entropy of physiological data itself.

The result shows two groups of subjects, one not showing much evidence of stress and the other exhibiting sufficient stress. Among the subjects who showed sufficient driving, 9 out of the top 10 high EEG-entropy drivers were female, one driver showed a strong pattern of range anxiety, and several showed patterns of uphill climbing anxiety.

SELECTED REFERENCES

Seyun Kim, W. Rhee, D. Choi, Y.J. Jang, Yoonjin Yoon, "Characterizing Driver Stress Using Physiological and Operational Data from Real-world Electric Vehicle Driving Experiment", 19(5), 895-906, International Journal of Automotive Technology, 2018