Human-robot interaction design for low cognitive load in cooperative work

Abstract

This work describes effective interaction design for a cooperative task in which a mobile robot sweeps over a table and a human can help it by picking up an obstacle. For efficient cooperation, we introduce behaviors of a robot to sweep a region under an obstacle when a human picks it up vertically, and designs the robot based on the interaction embedded in a human's operations to achieve a task. This interaction enables a human to easily cooperate with a robot without additional and explicit commands. We made experiments to evaluate the interaction embedded in a task by comparing with other explicit interaction, and the experimental results show that our cooperative sweeping is more efficient than autonomous one and our interaction embedded in a task needs less cognitive load than other explicit interaction such as voice or leading the robot by hand.