Marc Micatka

University of Washington Applied Physics Laboratory

Papers

Total Citations

H-Index

About

Marc Micatka is a robotics researcher focused on advancing autonomous manipulation in challenging, unstructured environments. His primary research areas lie at the intersection of robotic grasping, perception under uncertainty, and decision-making for manipulation in degraded or adversarial conditions. Micatka’s major contributions include pioneering methods for quantifying perceptual uncertainty in 3D reconstruction, particularly for underwater settings where sensory data is inherently imperfect. His work on "PUGS: Perceptual Uncertainty for Grasp Selection in Underwater Environments" introduces a novel framework that enables robots to account for and represent this uncertainty, leading to more robust grasp selection when visual information is incomplete. Additionally, his research on "Clustered Grasp Volumes for Improved Grasp Selection" addresses the critical challenge of identifying achievable grasps in unstructured scenes, offering a new approach to selecting grasps under adversarial environmental conditions. Though early in his career, with these foundational papers already garnering citations, Micatka is establishing himself as a key voice in perception-driven manipulation. His work is particularly notable for bridging the gap between theoretical uncertainty quantification and practical robotic grasping, making his research highly relevant for students and engineers working on field robotics, autonomous underwater vehicles, and manipulation in the wild.