A Wearable Interface for Intuitive Control of Robotic Manipulators Without User Training

Abstract

This paper describes the development of a wearable interface that exploits the user’s natural arm movements to intuitively control a robotic manipulator. The design is intended to alleviate the time and effort spent in operating the robotic manipulator, regardless of the age and technological experience of the user. The interface is made to be low-cost, comfortably worn, and easy to put on and remove. Kinematic models of human and robot arms are used to produce a natural mapping from the user’s arm movements to the commanded movements of the robotic manipulator. An experiment is conducted with 30 participants of varied ages and experience to assess the usability of the wearable interface. Each of the participants is assigned to perform a pick and place task using two of three different interfaces (the wearable interface, a game controller, and a mobile interface running on a tablet computer) for a total of 60 trials. The results of the study show that the wearable interface is easier to learn compared to the alternative interfaces and is chosen as the preferred interface by the participants. Performance data shows that the users complete the pick and place task faster with the wearable interface than with the alternative interfaces.