Point-and-command paradigm for interaction with assistive robots

Abstract

This paper presents a multi-modal interface for the interaction between a human user and an assistive humanoid robot. The interaction is performed through gestural commands and a dialogue mechanism to provide a 'natural' means for the user to command an assistive robot to perform several given tasks. A "Point-and-Command" interaction concept was proposed with the aim of providing the user with an easy-to-use and intuitive interface and to improve the efficiency of human-robot collaboration. Likewise, a decision support system (DSS) was implemented based on fuzzy logic, in order to warn the user when an active robot is about to become unusable and should be replaced. Thus, the interaction time, which depends on the robot's battery life, can be increased by replacing the assistant with another more fully charged one, making the interaction seem more natural.