A series of driving experiments were performed to detect physiological stress in a natural situation in which ambulatory artifacts were limited. The car testbed overcomes many of the difficulties encountered in the ambulatory experiments. The driving task limits the range of the subject's movements and creates a situation in which stressors are frequently encountered. The physical structure of the car permits placing video cameras at multiple locations. The stressors for the driving task were spontaneous, occurring naturally from driving on public roads. A particular driving route was specified to better control the types of driving situations encountered, taking the driver through resting, city driving, highway driving and toll situations. While driving, subjects were monitored with four types of physiological sensors: an electrocardiogram (EKG) to measure heart activity, an electromyogram (EMG) to measure muscle activity in the right trapezius muscle, a respiration sensor, and two skin conductivity sensors (GSR) placed on both the right hand and the left foot. A car based computer system sampled the signals and stored the digital data. Features were extracted from the data and tested in the stress detection algorithm. Three video cameras were used to continuously capture the driver's facial expression, road conditions and some body motion. An audio channel captured ambient noise and the driver's voice. The video and audio data were coded and used for labeling and validation.
For more detailed information, my doctoral thesis is available online, as well as a short paper from the ICPR '00 Proceedings through these links to Publications on the Media Lab's Tech Reports Page:
Wearable and Automotive Systems for the Recognition of Affect from Physiology (TR#526)
Jennifer Healey (June 2000)
MIT PhD Thesis - Electrical Engineering and Computer Science Department, June 2000.
SmartCar:Detecting Driver Stress(TR#525)
Jennifer Healey and Rosalind Picard (May 2000)
Appears in: Proceedings of the 15th International Conference on Pattern Recognition, Barcelona, Spain Sept 3-8 2000.