Publication
Toward Adaptive Robot Behavior for Interdependent Human-Robot Teams
Emilia Pietras; Stephanie Hall; Bernd Kiefer; Jeppe Langaa; Guglielmo Borzone; Mandeep Dhanda; Aljaz Kramberger; Vimal Dhokia; Norbert Krüger; Leon Bodenhagen
In: Proceedings of ICRA 2024. ICRA Workshop Dynamic Duos: Human-Robot Co-Worker Adaptation in Manufacturing, located at ICRA-2024, May 13, Yokohama, Japan, IEEE, 2024.
Abstract
In manufacturing today, a gap persists in the implementation of safe and efficient human-robot teams. To implement robot systems that truly collaborate with human teammates,
such systems must facilitate adaptive high-level robot action planning. This adaptive robot behavior should both account for the goal of the collaborative task and the characteristics
of their human teammate. Adapting to the human would not only take into account interpersonal differences but also the changes in the human’s physical and mental state throughout
the collaboration. In this paper, we will first present the related work in Sec. II, followed by our proposed model in Sec. III, which describes the modules needed to implement adaptive human-robot interaction (HRI) systems. The model consists of three main elements: 1) the human-robot team, 2) the knowledge base, and 3) the high-level action planning. These main parts each contain two or more subelements, which will be described in further detail. In Sec. IV, the method is put in the context of three different use cases: learning from virtual demonstrations, assembling and fixing components, and object scanning.