Feasibility study to distinguish movements automatically by analysing the pressure distribution on a seat

Abstract

Fatigue is an important factor in the occurrence of car accidents. In Germany, there are an average of almost 2000 car accidents with personal injuries due to tiredness of the driver per year [1]. Automatic detection of fatigue by constantly monitoring a person's condition allows the initiation of emergency braking and therefore reduces the number of car accidents. In this work, a pressure mat was used to record the movements of a driver, simulated by a male healthy volunteer. 18 sitting positions were defined and performed by the volunteer. In total, 103 measurements were evaluated. The results show, that it is feasible to detect movements, when the torso is moving. Movements of the arms without moving the torso were not clearly detectable. However, small differences in the quantitative measurements were detected. Using innovative artificial intelligence algorithms might enable the classification even if there is no torso movement included.