Using low-cost, wearable, sensors to detect driver state: Towards development of new algorithms for real-time detection of driver fatigue

Driver drowsiness is one of the leading causes of road-based crashes. Using a driving simulator study, I plan to help with the collection of data on drowsy drivers, analysing drivers’ physiological metrics, such as eye-tracking, heart rate variability and skin conductance, using low-cost consumer grade sensors. Using sensor fusion of the physiological data, I will co-develop machine learning algorithms that can detect driver drowsiness in real-time. This type of intervention can be useful in , providing drivers with warnings about their impairment, prompting them to take a break, to reduce the likelihood of crashes.

Faculty Supervisor:

Birsen Donmez

Student:

Partner:

University of Leeds

Discipline:

Engineering

Sector:

Education

University:

University of Toronto

Program: