Assessment of Walker-assisted Human Interaction from LRF and Wearable Wireless Inertial Sensors

Abstract

This paper describes the assessment of basic walker-assisted human interaction based on a laser range finder (LRF) sensor and two inertial wearable sensors. Thirteen osteoarthritis patients and thirteen healthy subjects were selected to be part of this pilot experiment, which intends to acquire and calculate spatiotemporal and human-interaction parameters from walker-assisted ambulation. A comparison is made between the spatiotemporal parameters of healthy subjects and the ones of patients with osteoarthritis. Moreover, it is made an analysis of the effect that change of direction in walker-assisted ambulation can have on spatiotemporal parameters. Results have shown that 1) velocity, step length and distance to the walker are significantly affected by the change of direction, and 2) distance to the walker and step length can distinguish between healthy subjects and patients with osteoarthritis. In terms of human-interaction parameters, results show that a LRF sensor can correctly describe the trajectory and velocity of the user in relation to the walker. However, just the wearable sensors can characterize changes in direction. These results will be further used in the development of a robotic control that intends to detect the user's intention through LRF and inertial sensors, and respond accordingly.