Content area
Multi-robot systems are increasingly being used to perform various and complex tasks, where robots are expected to collaborate with each other and interact with humans in dynamic environments. However, compared to single-robot interactions, interactions with multiple robots can complicate the way people perceive and interact with robots. Specifically, the flexible configurations of robot identities and embodiments in multi-robot systems can make it challenging for users to easily make sense of and manage interactions with such systems. Thus, this dissertation aimed at understanding users' mental models of flexibly configurable multi-robot systems and informing the design of the user interfaces used to manage interaction with multiple robots.
This work begins by discussing how different research methodologies in the field of Human-Robot Interaction can lead research participants to differentially make sense of experimental contexts and the robots involved in those contexts. In turn, a new theory is introduced, suggesting that certain methods may be better suited for investigating certain phenomena in Human-Robot Interaction---such as that of flexibly configurable identity in multi-robot systems.
This dissertation then investigates how different identity configurations of multi-robot systems influence users' mental models. In particular, it examines how users make sense of robots and what robot identity-related cues influence that sense-making. Additionally, it examines how those cues and the resulting mental models affect users' perceptions of entitativity and trustworthiness in multi-robot systems.
Next, this work demonstrates how flexibly configurable multi-robot systems may be represented within the algorithms and user interfaces used to manage multi-robot interactions. Specifically, this work provides a proof-of-concept of an user interface used to manage multiple robot identities and embodiments in social human-robot interactions.
Finally, this dissertation investigates the scope of user challenges and needs during the management of robot interactions. It explores how user interfaces can be designed to support users in the management of flexibly configurable multi-robot systems. This includes the introduction of a new visualization for tracking the capabilities of multi-robot systems and design recommendations for future user interfaces.
Overall, this work offers valuable insights on human interactions with multiple robots, particularly regarding users' mental models and management of flexibly configurable multi-robot systems.