Abstract: Informative Path Planning (IPP) is an important planning paradigm for various real-world robotic applications such as wildfire monitoring and predicting infection spread in crops. IPP involves planning a path that can learn an accurate belief of the quantity of interest, while adhering to planning constraints. Traditional IPP methods are effective only in static, time-invariant [...]

Abstract: Imitation learning enables scalable transfer of complex manipulation skills to robots, but its effectiveness depends on high-quality demonstrations and robust policy learning, especially in dynamic, contact-rich environments. This thesis investigates how combining imitation learning with teleoperation and classical planners can teach dexterous manipulation across diverse real-world settings. We develop a teleoperation system for collecting [...]

Abstract: Deploying AI in high-stakes settings requires models that are not only accurate but also interpretable and amenable to human oversight. Concept Bottleneck Models (CBMs) support these goals by structuring predictions around human-understandable concepts, enabling interpretability and post-hoc human intervenability. However, CBMs rely on a ‘complete’ concept set, requiring practitioners to define and label enough concepts [...]

Abstract: Generalization is a core challenge in robotics, where the goal is to enable robots to handle novel objects, environments, and embodiments with minimal additional data. This thesis explores how human prior knowledge, captured through both passive observation and active demonstration, can be leveraged to improve generalization in manipulation tasks. We propose two complementary approaches that scale robot learning leveraging large-scale human-derived data. First, we introduce HRP (Human Affordances for Robotic Pre-Training), where we learn actionable visual representations by extracting hand trajectories, contact points, and object labels from internet-scale human videos. These representations, when used to initialize control policies, lead to significant performance gains in downstream robot manipulation tasks and transfer effectively across viewpoints and robot morphologies. Second, we present DexWild (Dexterous Human Interactions for In-the-Wild Robot Policies), a system that collects high-fidelity in-the-wild demonstrations using a human motion-capture device. A human-robot co-training algorithm combines this diverse human data with limited robot data, enabling robust policy transfer to unseen scenes, robot arms, and hands. [...]

Abstract: As robots and intelligent systems increasingly interact with humans, the ability to understand users by anticipating their actions becomes significantly more important. Current approaches to action anticipation leverage the inference capabilities of large foundational models but are limited in their application by the complexity and resource requirement, as well as the difficulty of training. [...]

Abstract: Assistive robotic systems have the potential to significantly enhance autonomy and independence for individuals with physical disabilities. However, existing shared autonomy frameworks typically rely on static policies trained offline, which fail to adapt effectively when encountering unforeseen environmental variations or evolving user behaviors. Additionally, teleoperating high degrees-of-freedom (DoF) robotic manipulators through low-dimensional (low-DoF) user [...]