Moving Together: A Performing Arts Approach to Relational Human-Robot Interaction Design

This dissertation investigates novel perspectives on movement-based human- robot interaction (HRI), integrating insights from the performing arts, cybernetics, phenomenology, and cognitive studies to develop a relational framework for HRI design. It first demonstrates that the manner in which movement-based HRI has been conducted and theorised has overlooked aspects of interaction and communication that are important for creating and sustaining meaningful relationships among humans and robots. Then, it develops a conceptual framework by drawing from the knowledge and practice of the performing arts. This dissertation introduces the conceptual pair of the internalist and relational views of interaction as a lens to assess design choices in the field of HRI. The internalist view focuses on the encoding and decoding of internal states (emotions, beliefs, intentions) and dispositions (traits, tendencies) that drive behaviour. In this perspective, movement is seen as a medium to express and reveal these hidden, predetermined, and relatively universal internal states, making them observable to the interacting partner. This approach is prevalent in human-human interaction (HHI) and HRI, where the goal is often to decode these internal states uniformly to predict and understand behaviour. In HRI and related fields such as cognitive studies and psychology, alternative views are emerging that adopt a more relational perspective on interaction, highlighting how human-robot encounters are deeply contextual and situated, always part of larger dynamics. They also consider movement as more than just a medium for conveying symbolic messages, viewing it instead as a corporeal tool for creating dynamics and patterns among interactants. Finally, these views propose that the states exchanged during an event do not pre-exist the encounter but are co-created by the interactants through their engagement. While the internalist view is more established and prevalent in human- human interaction (HHI) and HRI discourses, the relational lens remains relatively underrepresented. This dissertation aims to develop the relational view further conceptually and practically through my own empirical work, highlighting how movement can foster meaningful connections between humans and robots beyond its potential to convey internal states and dispositions. My goal is to intervene in the field of HRI by providing additional tools for researchers and designers. These tools include conceptual resources to help navigate the field of HRI and critically reflect on its underlying assumptions, as well as theoretical and practical recommendations to approach HRI design from a relational standpoint. In order to develop this relational framework, this thesis uses the knowledge and practice of the performing arts in three main ways. Firstly, I draw on theoretical insights from the field of performance studies that emphasise relationality and performativity. Secondly, I turn to the practice of robotic art as an interdisciplinary art form that can inspire novel HRI designs. Finally, I rely on the embodied expertise of performing art professionals in situated and improvised scenarios with robots. Here, the corporeal knowledge of these experts is leveraged to enact encounters with robots in improvised settings, from which this thesis derives design recommendations.