A Proposal of Speech- Driven Entrainment-Feedback Towards Smart Vehicle System

抄録

Smart electric vehicles (SEVs) with self-driving ability are likely to become a common technology in near future therefore drivers are most likely very relaxed and become inattentively participate in driving processes. However, by viewing the SEVs solely from their electromechanical computing technology perspective are not holistic enough to encapsulate them as ideal smart vehicles. In other words, the SEVs with self-reliance on driving technology alone is not good enough without pertinent involvement of the drivers. In this paper driver's speech entrainment communication between a Line Follower Mobile Robot (LFMR) was evaluated and established connection between driver's mental states and performance of the LFMR. A speech recognition system evaluated driver's speech and classified them into three different mental states which were angry, anxiety and neutral then they were manually embedded into a Basic Stamp Microcontroller (BSM) on the LFMR. The BSM randomly combined two mental states in a set of serial sequences and translated into autonomous driving performance. Finally the performance was measured based on speeds, accelerations and travel periods. The LFMR attained the highest average speeds and shortest time travels whenever angry mental state was present meanwhile anxiety produced the poorest results in all mental state combinations.