An argumentation-based dialogue system for human-robot collaboration

Abstract

Human-robot collaboration may fail due to conflicts in beliefs or plans between cooperating partners, or due to robot errors. Dialogue is an intuitive way to resolve such conflicts due to miscommunication. The research demonstrated here explores the notion of using argumentation-based dialogue for human-robot interaction. The demonstration presents a proof-of-concept prototype of a logic-based dialogue framework grounded in argumentation theory that addresses the what to say problem in human-robot communication during a collaborative task. A simulated human-robot treasure hunt game is shown, where a robot searches for objects of interest in a region that is not accessible to a human and interacts with the human in order to interpret its sensor data and complete the task effectively.