Facilitating human-robot interaction: A formal logic for task description

Abstract

We develop a formal logic for task descriptions that are easy to interpret for both humans and robots. Tasks are described in propositional forms that reflect syntactic structures observed in natural language so that all the resulting task descriptions can be easily understood by humans. At the same time, these propositional forms ensure that each task description is interpretable by robots. Infinitely many task descriptions can be created in this formal logic so that the formal logic can support complex human-robot interactions. We establish a hierarchy of propositions that enhances the expressive power, the interactivity, and the deductive apparatus of our formal logic. We also examine how to systematically evaluate the feasibility of each task description using the formal logic.