According to the latest statistics by the National Highway Safety Administration, approximately 1,830 young drivers (15–20 age) are involved in fatal crashes. Driving tasks such as keeping the vehicle within the lane, observing traffic signs, turning, and identifying any sudden danger involve various cognitive processes including attention, memory, vision, spatial orientation, motor control, executive function, and decision making. This study proposes an innovative way to assess driver’s behavior, especially focusing on young drivers, by applying the emerging sensing technologies, functional near infrared spectroscopy (fNIRS), driving simulator, and analytical methods (e.g., machine learning). Study findings show machine learning can achieve an average of 97.5% accuracy when using fNIRS data and 70% accuracy when using driving data. While the team acknowledges the limited data set is used in this paper, this study opens the doors for applying the emerging machine learning methods to driver’s cognitive processes to further understand young individual’s driving behavior.
Understanding Driving Behavior Using fNIRS and Machine Learning
International Conference on Transportation and Development 2021 ; 2021 ; Virtual Conference
2021-06-04
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Analysis of braking intention based on fNIRS in driving simulation experiments
Wiley | 2019
|British Library Conference Proceedings | 2008
|Analysis of braking intention based on fNIRS in driving simulation experiments
IET | 2019
|