Driver physiological monitoring for vehicle Emergency Response (DrivER)

Important results from the project

The aim was to develop concepts for detecting driver states: 1) Sleepiness detection in real driving; 2) Assessment of stress level in professional drivers using discrete sensors. 3) Detection of vitals and sudden illness due to cardiac complications such as arrhythmias. The project has generated new knowledge and IP for all concepts, the results show that detecting driver states and sudden illness with non-invasive sensors in a vehicle environment is challenging, but COPE DrivER has developed methods to tackle the challenges that warrant further development.

Expected long term effects

The knowledge building and the network the project has built has led to further ongoing projects and additional projects that have been defined. An example of an ongoing project is Syncope, which continues to investigate the possibilities of detecting sudden illness. Autoliv continues data collection for sleepiness detection. The project has currently resulted in 4 peer-reviewed scientific publications and one submitted. Expected effects are that efforts to develop the technologies towards implementation increase.

Approach and implementation

Experiments were designed and conducted to complete data collection through driving simulator (for stress and mental load), naturalistic driving (for detection of sleepiness through heart rate measurement and machine learning methodology), as well as experiments in a lab environment at Chalmers and Raytelligence. Statistical tests and different strategies for training and evaluation of machine learning models were used, which were compared with the research subjects´ perceived stress and sleepiness level and against reference equipment.