An Approach For Robots To Deal With Objects

Abstract

Understanding object and its context are very important for robots when dealing with objects for completion of a mission. In this paper, an Affordance-based Ontology (ABO) is proposed for easy robot dealing with substantive and non-substantive objects. An ABO is a machine-understandable representation of objects and their relationships by what it's related to and how it's related. By using ABO, when dealing with a substantive object, robots can understand the representation of its object and its relation with other non-substantive objects. When the substantive object is not available, the robots have the understanding ability, in term of objects and their functions to select a non substantive object in order to complete the mission, such as giving raincoat or hat instead of getting stuck due to the unavailability of substantive object, e.g. umbrella. The experiment is done in the Ubiquitous Robotics Technology (u-RT) Space of National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Japan.