Junior Associate Professor
Department of Applied Computer Engineering,
School of Information Science and Technology,
We have been studying “active bone-conducted sound sensing”, which can estimate joint angle. Joint angle is estimated by sensing the variation of the transmitted vibration input to the bone. We have already actual experiences on elbows and finger joints, and we believe it will be usable also to other joints. Additionally, we have studied “passive bone-conducted sound sensing,” which can recognize a tapped position. If you are interested in these researches, please contact us.
We input vibration actively to a bone by an actuator and sense the variation of the transmitted vibration in order to estimate the joint angle. Since the frequency of the vibration is higher than user senses it, it does not cause user uneasiness. Although there are differences between individuals, once the model is gained, it enables us to estimate joint angle. Unlike other methods to require relative angle variations, the main advantage is to be able to estimate the absolute angle. Since frequencies of bone-conducted sound by tapping and from actuator are different, active and passive bone-conducted sound sensing can be used simultaneously.