Self-driving cars is considered the next major step in the automotive industry and with automation in passenger vehicles, the driver can benefit from the freed up time for leisure or work, as he or she becomes the passenger. However, this is only possible if the drivers are comfortable during automated driving. The major issue here, is that the susceptibility of motion sickness (MS) increases significantly when the driver does not have his or her eyes on road, this susceptibility needs to be minimized. However, motion sickness is not clearly understood and assessing it has been traditionally done qualitatively with questionnaires. Because of the highly individual and subjective nature of motion sickness and its symptoms, it is hard to quantify it accurately by questionnaires. To minimize and prevent motion sickness, it is beneficial to measure it quantitatively in different ways in addition to the traditional questionnaire.
This thesis explores and investigates the use of physiological measurements, ECG and GSR, to relate motion sickness in a realistic driving experiment.
Hence, a road test is conducted for this study with a Toyota Prius on a closed road. A slalom course was driven with a speed 25 km/h to reach lateral accelerations up to 0.4G with a lateral frequency of 0.175 Hz. This frequency and velocity has been chosen, because it is known that people are the most sensitive for MS of frequencies near 0.2 Hz and this velocity reflects urban driving. 23 participants took part of the experiment and had their ECG, GSR and their MISC (Misery scale, an illness rating) recorded during the drive. The experiment lasted until MISC rating 7 (=medium nausea) or either 30 minutes was completed. The ECG (HR, LF/HF ratio) and GSR (skin conductance level SCL, skin conductance response SCR) recordings were then compared to the MISC ratings to see if there was a significant difference between the participants who got sick and stopped at MISC 7 (sensitive) and the participants who did not get sick (non-sensitive).
The results show little support correlating motion sickness or even to distinguish sensitive and non-sensitive groups with HR, HRV or GSR data. It might be beneficial to categorize people into different sensitivity profiles for MS susceptibility to make HR information more useful as there is too much of individual differences. Currently, there were few to none road tests done regarding motion sickness. It appears that physiological measurements for predicting MS in vehicles are not as straightforward and do not translate well from other types of laboratory tests to realistic road tests. Not to mention that HRV has become a controversial metric in the recent decade that might need to be re-evaluated for use.