Monitoring the driver

The needed information from the driver is the driver condition like the stress level, the fatigue level, the alcohol level and the distraction of the driver. The stress level can be measured using an ear sensor that provides information about the heart rate. The heart rate can be used to derive the heart rate variability (HRV). The HRV is the variation of the interval between two heart beats, also called inter-beat interval(IBI) and allows to detect the stress level of the driver [65, 66, 67], as high stress leads to a high HRV. However, a low HRV indicates a low stress level. The stress level of the driver can be gathered by the driving system either by using a software that calculates the stress level and passes the information to the interface module or by attaching a sensor to the interface module and calculating the driver stress level on the basis of the HRV.

According to Jung et al. [68] the analysis of the HRV also gives a valuable informa- tion about the fatigue and drowsiness status of the driver. They calculated the HRV of the driver on the basis of the ECG signal and analysed the power spectral density distribution of the HRV across its very low- (0.003-0.04 Hz), low- (0.04-0.15 Hz) and high-frequency. Jung et al. showed that the ratio from the low- to high-frequency can be used to decide if the driver is in normal, drowsy or fatigue condition, as the ratio from the low- to high-frequency decreases when the condition of the driver pro- gresses from awake to drowsy [69]. Another way to detect drowsiness and fatigue is by tracking the eyes of the driver. Singh et al. [46] proposed a driver fatigue monitor- ing system that monitors the driver eyes using a camera and warns the driver when the driver shows symptoms of fatigue. The driving system of Daimler [70], Attention Assist, observers the driving behaviour of the driver. On detection of the typical indicators of drowsiness in the steering behaviour, a warning is given to the driver. Based on the presented fatigue or drowsiness detection systems, the fatigue or drowsi- ness information can be gathered by using the in-vehicle serial-bus system when the fatigue or drowsiness detection systems are connected to the car or by attaching such a fatigue or drowsiness detection system directly to the interface module.

There are four kind of sensors for measuring the blood alcohol level [71, 72, 73, 54]: skin sensors, sweat sensors, alcohol sniffers and breath sensors. There are already skin sensors developed [74, 75], also called tissue spectrometry systems, that need skin contact to determine the blood alcohol level. According to the American Beverage Institute [71] and USA Today [76], Toyota started to develop a steering wheel in 2007

3.6. INTERFACE MODULE

that embedded such a skin sensor to detect the blood alcohol level. Sweat sensors, also called transdermal sensors, need also skin contact to detect the blood alcohol level of the driver. Nissan [77] presented in 2007 a concept car that integrated a sweat sensor in the gear shift knob to prevent drink driving by blocking the transmission of the car. Breath sensors are widely used for estimating the blood alcohol [72]. The sensor measures the concentration of alcohol in the breath, as ethanol is able to partition itself from the capillary blood into the inspired air. Volvo [78] presented in 2007 a driving system that is using a breath sensor to prevent drink driving by prohibiting to start the engine on detection of alcohol. Alcohol sniffers, also called distant spectrometry systems, are trying to detect the presence of alcohol in the air. According to Ferguson et al. [73], alcohol sniffers are trying to detect the alcohol concentration of the breath for example within the driver cabin without the need to provide a deep-lung breath like in breath sensors. Such sensors can detect alcohol also when the window of the car is opened and the air conditioner of the car is set to recycle [54]. Nissan [77] integrated, additionally to the sweat sensor, an alcohol sniffer in the seat of their concept car in 2007 to detect the blood alcohol of the driver. On the basis of the presented sensors and alcohol detection systems, the alcohol level of the driver can be gathered either by connecting the sensors directly to the interface module or to the in-vehicle serial-bus system for example by embedding the sensors in to the driving wheel or driving seat, as shown in the concept car of Nissan [77].

The visual distraction and the cognitive distraction are combined in order to get the distraction level of the driver. The visual distraction can be measured for example using an eye tracking system. Volvo [79] developed a head and gaze tracker, called Volvo/ANU system, that is able to robustly track head pose, gaze and eye closure in real-time in the environment of a car. Cognitive distraction can be measured by using the physiological signals of the driver [80] or the driving performance. However, according to Lee et al. [80], the relationship between the physiological measures and the driving performance might be a particularly powerful predictor of distraction such as the eye movement and steering behaviour. Thus, a combination of sensors and driving performance measures can be used for detecting the cognitive distraction of the driver, whereas a stereo camera can be used for detecting the visual distraction. The sensors and cameras for detecting the distraction level can be connected to an ECU of the in-car serial-bus system that calculates and provides the distraction level to the interface module or by connecting the sensors and cameras directly to the interface module and calculating the distraction level within the driving system.