A study on human-robot collaboration for table-setting task

Abstract

As the robot technology is advancing, it is possible to use robots for basic day-to-day chores, so that the burden can be taken off from humans. To make the robots perform such tasks, it is necessary for them to handle different types of objects. Manipulation of deformable objects such as cloth is a challenging task for a robot because of high dimensionality and large number of possible configurations of cloth. Previous studies have covered large number of simple manipulations of cloth articles. In this paper, we are focusing on table setting task that requires putting on a sheet of cloth on the table. This paper proposes human-robot collaboration for table-setting task based on visual assessment. We used Baxter robot to hold two corners of rectangular tablecloth and other two corners are held by human. A head-mounted Kinect sensor is used to get the state of cloth and Robot arms are used for controlling the position of cloth corners. We use features from Kinect sensor to assess whether the placement of the cloth is successful or not. We demonstrate an initial study of the system that can achieve promising results towards table setting task through human-robot interaction.