Dongwon Son

Seoul National University, Samsung (South Korea)

Papers

Total Citations

H-Index

About

Dongwon Son is a robotics researcher whose work pushes the boundaries of large-scale, dexterous robotic systems and intelligent grasping. His primary research areas include aerial manipulation, robotic grasping in cluttered environments, and data-driven simulation for contact-rich tasks. Son’s most influential contribution is the development of **LASDRA** (Large-Size Aerial Skeleton System with Distributed Rotor Actuation), a novel framework that overcomes the limitations of traditional internal actuation by distributing rotors across a large, articulated aerial structure. This work, which has garnered **53 citations**, opens new possibilities for constructing large, dexterous robots capable of handling significant external forces. In grasping, Son introduced a reactive, closed-loop system that predicts 6-DoF grasps in heavily cluttered settings, treating grasping as an inference problem—a notable advance over conventional 4-DoF top-down approaches. His work on data-driven contact clustering further enhances the fidelity of rigid-body robot simulations by learning contact patterns from real sensor data. Through these contributions, Son is shaping the future of autonomous manipulation and large-scale robotic systems.