Improved usability of locomotion devices using human-centric taxonomy

Abstract

This thesis investigates the failure of early taxonomies for locomotion to provide a comprehensive enough framework in facilitating the development of usable locomotion devices due to an inadequate classification of the human component. It then proposes an alternative human-centric taxonomy for locomotion that grounds itself on the physiological, physical and extra-physical cues the human body is capable of providing rather than only the input existing interaction devices are capable of receiving. Through the realization that interaction begins with the human, not the machine, this thesis is able to determine a cue from the body that is able to provide enough information for use by an algorithm to recognize walking and running forward, sidestepping, back stepping, and jumping with a minimal amount of input. This thesis then develops and performs initial tests on a fully implemented locomotion device using input from two inertial sensors on the legs in conjunction with the locomotion recognition algorithm for use in any commercial-off-the-shelf (COTS) video game for PCs that use keypresses for locomotion input.