Inclusive human-robot interaction for gait rehabilitation and wheel-chair exercises

Abstract

The paper describes inclusive human-robot interaction for gait rehabilitation and wheel-chair exercises. Robotic interaction is a promising method for prospective gait rehabilitation. In literature, gait training platforms such as power assisting limbs and body supporting platforms have been studied well. These types of platforms mainly assist physical movements of limbs and body balance. In these approaches, however, cognitive assistance and its effect for rehabilitation are not enough discussed. In the proposed framework, we focused on subject's own motor recognition and the effect for recovery of motor functions. We developed an inclusively interacting robot, Lucia that evaluates gait patterns of patients and navigates the exercise by audio, visual and somatosensory stimulation. In this paper, we will propose our latest algorithms of gait measurement and introduce a newly developed robot hardware. We performed experiments and evaluated accuracy of leg tracking and landing detection. The experimental results showed sufficient accuracy for gait measurement. We also applied the tracking algorithm for inclusive interaction with wheel chair users. In the interaction, we observed that the wheel chair users enjoyed playing with autonomously following Lucia. The interaction was safely demonstrated by whole body touch detection and easy stop control even by wheel chair users.