Alexander Lambert

University of Washington, Georgia Institute of Technology, Nvidia (United States)

Papers

Total Citations

H-Index

About

Alexander Lambert’s research lies at the intersection of robot motion planning, tactile sensing, and state estimation, with a focus on enabling robots to operate safely and intelligently in complex, contact-rich environments. His most influential work, “Stein Variational Probabilistic Roadmaps” (2022, 5 citations), introduces a novel sampling-based motion planning framework that leverages Stein variational inference to efficiently generate high-quality planning graphs, directly addressing the critical need for reliable global path plans in autonomous systems. In the domain of physical interaction, Lambert has made significant contributions to robust state estimation. His 2019 paper “Joint Inference of Kinematic and Force Trajectories with Visuo-Tactile Sensing” (3 citations) pioneers a method for simultaneously estimating a robot’s motion and the forces it experiences during manipulation, a key challenge for dexterous tasks. Complementing this, his work “Robust Learning of Tactile Force Estimation through Robot Interaction” (3 citations) develops learned models that accurately map tactile sensor signals to force, overcoming the limitations of inaccurate analytic models. Earlier, his 2018 work on deep forward and inverse perceptual models (2 citations) advanced the ability to predict visual outcomes from robot states and infer states from images. Through these contributions, Lambert is shaping the future of autonomous manipulation and safe robot deployment.