Emotional Reliability Design: Human Driven and System Supported Approach

Traditional system reliability approach is to design a system which will satisfy required reliability for a fixed mission. The mission does not change and the major role of a human is to respond properly when an emergency takes place and to take necessary actions to get the system back to normal. But a human gets panicked at the time of emergency and many human errors have been reported in such situations. Recent brain study reveals that human capabilities will decrease to 30 % of their full power. Now situations are changing very frequently and extensively. Understanding the situation is becoming very important. The problem with the traditional approach is that it was assumed that a designer can foresee the situations well beforehand. But it becomes extremely difficult now for a designer to foresee the changing operational situations. Therefore it become more important for a system to help a user perceive and understand the situations and make adequate judgment and take necessary action. Why we cannot leave too much to a user is because when something unexpected happens, he may get panicked and become emotional unstable and he cannot act as expected. If we can bring him back to his normal emotional state, he could understand the situation better and cope with it more reasonably, flexibly and adaptively. This paper describes our attempts to detect a car driver´s emotion with the aim of developing a system to keep him emotionally stable for safer driving.