A face vector - the point instruction-type interface for manipulation of an extended body in dual-task situations

A human's work efficiency can be improved in many situations if performing two different tasks simultaneously is possible. In recent years, various robotic extended bodies and its control interfaces have been proposed for assisting manual work. Most of these interfaces are path control-type operation and have been verified under single-task situations; however, their usability with respect to voluntariness and intuitiveness in a dual-task situation has not been addressed. In this paper, we focus on point instruction-type operation that can be used together with conventional path control-type operation in order to improve the control intuitiveness, and propose a new interface 'face vector' which can point at 3-dimensional target position in non-task-oriented environment. The voluntariness and intuitiveness of proposed modality is discussed by measuring its pointing accuracy and effect on work efficiency in dual-task conditions. The results show that numerical data of pointing accuracy of face vector in single and dual task conditions, and indicates that proposed interface is enough intuitive not to reduce the work efficiency.