Pierson, AlyssaCoffey, Mela2025-05-222025-05-222025https://hdl.handle.net/2144/504682025As robots continue to become more sophisticated, we need robots capable of interacting and collaborating with other robots and human teammates. We envision robots designed to collaborate with other agents, whether robot or human. This vision presents a variety of challenges, and this dissertation specifically focuses on exploring various levels of inter-agent interaction and inferred teammate information. We first explore single-human–single robot teaming within the scope of collaborative navigation, which requires direct human-robot interaction and inferred teammate information. We consider the case where a human and robot must navigate through an environment to reach a shared goal. We propose two distinct haptic guidance systems, which provide collision avoidance and route suggestions to the human user via force feedback. Each approach provides different benefits in computation and optimality. Analytical and experimental results show our approach guarantees simultaneous collision avoidance and guidance to the goal. While this first work addresses two-agent collaboration, the remainder of this dissertation focuses on multi-agent collaboration, with little interaction between robots and humans and different levels of teammate information. We consider a heterogeneous multi-agent team, in that each robot has different capabilities such as sensing or resource capacity. We propose control policies to deploy robots to serve complex, dynamic tasks, enabling the team to adapt to changes in tasks and within the team. Our distributed approach allows for large-scale applications. Analyses show our approaches converge to optimal solutions while adapting to the dynamic environment, and simulations and hardware experiments demonstrate comparable performance to baseline algorithms.en-USAttribution 4.0 Internationalhttp://creativecommons.org/licenses/by/4.0/RoboticsThesis/Dissertation2025-05-220000-0003-2275-8341