Emotional Profile Assessment for Safe Driving

1. BackgroundThe Internet of Things (IoT) aims to integrate technology into everyday lives. In a desirable IoT scenario, technology would take peoples context into account to simplify their lives and improve health and safety by offering services that work according to users emotions and preferences. Such services can be offered to users in line with the emerging platform which we called IoET. Safe driving is an application which is needed to be taken seriously due to life risks and can take the advantages of IoET. Strong emotions can be a source of distraction and a number of studies link highly aroused stress states with impaired decision making capabilities, decreased situational awareness and degraded performance which could impair driving ability [1]. Therefore, it is important while drivers experiencing strong emotions, a driver assistance system (DAS) notice them, before they decide to react on the road. 2. Method Based on psychophysiology findings, emotions and physiology are closely related. Hence, measurement of physiological parameters generated by the activation of the sympathetic nerves of Autonomous Nervous System (ANS) can be helpful to decide the emotive status of human being. In this way, to understand the drivers state for using in a DAS, physiological signals are useful metrics which collected by body mounted sensors without interfering with the drivers task performance. A project by Healey and Picard to determine a driver’s relative stress level by using ECG, EMG, skin conductance (SC) and respiration sensors shown in Fig1[2]. But among all of the measurable physiological signals, which are suggested for stress detection, we identified the SC as a better marker since it does not affect by users motions and activities. The Galvanic Skin Responsesensor (GSR) allows us to measure the stress level of users, hence to evaluate the measurement accuracy and to find an appropriate placement based on concentration of eccrine sweat gland [3] (Fig. 2) for unobtrusive data collection, we conduct a series of tests. We attached three sets of GSR sensors to left hand fingers, postauricular area and forehead area as parts of the lymphatic system and an Infiniti Biograghs SC sensor on right hand fingers as a gold standard and collect data while users participating in a standard stress test of Biograph system 3. ResultsIn this paper, we present methods and tools required to measure stress level and address challenges such as data collection, communication, and user experience in our implemented testbed. The results of our primary tests from 15 subjects over 14 minutes stress test have verified our collected data from GSR sensors and showthat the SC could be measured from post auricular area and forehead. A sample of the results is depicted in Fig. 3. In the full paper, we demonstrate how collected GSR data in a vehicle can lead to DAS to alert a driver. 4. Conclusions In general, the results suggest that stress level of users can be detected and measured using GSR biosensors placed over different body areas resulting different accuracy.