Assessment of driving skills using Virtual Reality: Comparative survey on experts and unskilled users of electric wheelchairs

Abstract

This survey addresses the issue of assessment of a person's driving skills to operate a powered wheelchair. Usually, the acquisition of a new wheelchair implies several evaluation tests to fit the wheelchair's configuration to the specificities of the user's disabilities. Usually, such tests are made in the actual world and based mainly on qualitative observations. Virtual Reality has been applied to transpose the actual tests into a Virtual World, and define methods completed with quantitative criteria to put into evidence the ability that a person has to drive an electric wheelchair. To do this we propose a comparative survey that implies tests made in a virtual world to investigate the differences between experienced users with disabilities, and unskilled able-bodied ones. The results of the tests allowed to define two new quantitative criteria related to the joystick position x_J (along its X axis). One is based on the amplitude of the x_J signal and the other on its Power Spectral Density. Although both of them contain valuable information, the association of the proposed criteria appeared to be more efficient to identify the skilled and unskilled subjects.