Abstract
First, we aim to realize effective human-robot interaction in two scenarios: a robot (1) recommends short breaks for a person performing lengthy computer work and (2) promotes light exercise while a person is watching television in a living room. We study detection methods for reduced concentration levels to determine the appropriate timing of the robot's interruptions. Our results confirmed that heart rate and respiratory signals vary with concentration, and non-intrusive measurements are required in such cases; therefore, we introduced a chair with sensors. We use capacity-coupled electrodes for ECG measurements, and respiration is measured by changes in the back pressure detected by sensors attached to air bags. Second, we develop tools that assist users to learn breathing methods that can relax their minds and bodies. To guide the respiratory rhythm, we propose three easy-to-use prototypes: (1) a chair-type tool that applies music and illusory sensations called “phantom sensations,” (2) an iPad app that measures respiration using an embedded accelerator and controls the tempo of music, and (3) a stuffed toy for children that guides their respiration by inflating its abdomen.
