We present a new technique that measures driver fatigue from the dynamics of driver-vehicle interfaces, such as the accelerator and the steering wheel. Real-life highway driving experiments were conducted on 24 participants and the variance or energy of the time series data on driver imparted force or motion on the interfaces was found to be strongly correlated to the driver's self-perceived alertness. The non-visual sensing principle and computational algorithm demonstrate the simplicity, robustness and cost efficiency of the proposed technique, which is expected to impact the transportation industry, particularly the automotive sector.
All Science Journal Classification (ASJC) codes
- Automotive Engineering
- Mechanical Engineering