Maintaining postural stability during walking is important, particularly for individuals with impaired balance control, such as the physically weak elderly. It is known that light touch results in reducing body sway and enhancing postural stability by providing additional somatosensory information. Therefore, robotic light touch is expected to stabilize the walking motion of individuals with impaired balance control. This paper proposes a three-dimensional lumbar trajectory prediction model considering human anatomy during walking for implementing future consistent robotic light touch at a fixed point on the user’s lower back with a constant contact force. The lumbar trajectory model is considered by combining a spatial pelvis motion model with the relative lumbar joint motion with respect to the pelvis. Additionally, the relationships between model parameters and walking speed are explored to enable subject- and speed-dependent trajectory prediction. The accuracy of the proposed model is evaluated using walking data from healthy young adults. The proposed model provides a promising foundation for enabling future robotic light-touch assistance systems throughout human walking.
Chen et al. (Wed,) studied this question.