A novel vision-based human-machine interface for a robotic walker framework

Abstract

This paper presents an innovative Human-Machine Interface (HMI) for a robotic walker. Robotic walkers offer their users an aid for sustaining mobility and the potential to rehabilitate their lower limbs. Mobility is a crucial function for a human being and these aids are paramount for improving the independence and quality of life of those who suffer from some form of mobility impairment. However, important factors like cost and safety make them either frequently inaccessible to the public or discarded due to a lack of confidence in their operation. The approach adopted in this work offers an intuitive human-machine interface combined with innovative safety measures. This result was possible due to the resourceful use of the low-cost Leap Motion sensor. Experimental evaluation was divided into two stages. First the system was tested using a simulated environment which served to validate the principle of operation of the HMI. In the second stage the proposed HMI was tested on board a robotic platform to evaluate its performance in a real-world scenario. Experiments performed with healthy volunteers revealed an intuitive user interaction and accurate user intention determination.