Date of Award
Master of Science
Driving is known as a daily stressor and measurement of driver's stress in real-time can improve the awareness of stress for drivers, their cars, and their phones. Integrating sensors in future cars can help assess driver's stress, but it requires either wearing sensors by the driver or instrumenting the car. In this thesis, we present "GStress", a model to estimate driver's stress using only Smartphone GPS traces. By obviating any burden on the driver or the car, our approach has a better chance of wider adoption worldwide. The GStress model is developed and evaluated from data collected in a mobile health user study where 10 participants wore physiological sensors for 7 days (for more than 10 hours) in their natural environment, including during driving. Each participant had 10 or more driving episodes over the course of the study (for a total of 37 hours of driving data). This being the first work of its kind, provides a correlation of over 0.7 between the actual and estimated driving stress by identifying some major factors such as stops, turns and brakings that contribute to the stress of a driver. Incorporation of other factors in the model as well as use of more advanced modeling approaches can further improve the accuracy of the model.
dissertation or thesis originally submitted to the local University of Memphis Electronic Theses & dissertation (ETD) Repository.
Vhaduri, Sudip, "Estimating Drivers Stress from GPS Traces" (2014). Electronic Theses and Dissertations. 877.