In this paper the challenges of developing a drowsiness detection system for heavy vehicles is described, but also the great need for such a system. The accidents have a very high fatality rate and are more common with commercial vehicles due to the nature of the work. Both the phenomenon drowsiness itself and the methods used for measurement are highly complex and subject of much research. The problem of an absent ground truth is presented and exemplified with two previous simulator studies. A more superior method of verification under development is presented at the end.
Abstract