Welcome to the TEA Lab
Welcome to the Thinking Embodied Agents (TEA) Lab at Virginia Tech! We build robots that operate in human environments rather than around them. We study how robots acquire and retain manipulation skills across their lifetime in unstructured settings, and how they model and adapt to the people working alongside them. We pair learned policies, such as language-conditioned, generative, and vision-language-driven approaches, with symbolic structure such as knowledge graphs, planners, and movement primitives, so that behavior generalizes from few demonstrations and stays inspectable enough to reason about safety and contact.