Development of a measurement system for train driver alertness detection

With the development of science and technology, people are paying more attention to traffic security issues. Drowsy driving is present in a large amount in causes of road traffic injuries, and its hazards become increasingly serious. Based on mechanical measured quantities, this thesis aims to develop a measurement method to detect train drivers’ movements. Nowadays, there have been lots of ways to identify drivers’ condition. In this thesis, we combine different sensors, especially using force sensors, strain gauges, IR detectors and webcams for train driver’s movements detection. Correlation was made between the information obtained from signals of the sensors and the monitoring video. The first phase of this work provides preliminary testing aimed to screening of the used sensors on the basis of the detection sensitivity. A second phase of the performed activity was related to the assessment of the sensors performances with respect to their efficiency for the reduction of the required feedbacks to the driver. As a result, an acceptable range of required feedbacks reduction for IR sensors and pressure maps was found. Moreover, with the developed system, it was evidenced that the number required feedbacks is strongly reduced by using more than one sensor.