Detection of Emergency Braking Intention Using Driver's Electroencephalographic Signals

This work investigates the recognition of the intention to perform emergency braking from driver's electroencephalographic (EEG) signals. To do so, brain signals and vehicle data were recorded in a simulated driving environment while participants had to drive and to avoid potential collisions by performing emergency braking. To resemble realistic conditions, emergency braking were performed during the presence and absence of stress, workload and fatigue. Brain signals were used to study the classification between emergency braking intention and normal driving. The results showed significant classification accuracies around 80% using EEG signals from the left hemisphere. On the basic of these results, this work shows the feasibility of incorporating recognizable driver's brain signals into advanced driver-assistance systems to carry out early detection of emergency braking situations.