Wenzhao Lian
Papers
10
Total Citations
173
H-Index
8
About
Wenzhao Lian is a robotics researcher specializing in robotic manipulation, cable handling, and robot learning, with a focus on bridging perception, planning, and dexterous control. His most recognized contributions center on robotic cable routing, where he has developed spatial representation frameworks and tactile-driven motion primitives to tackle the notoriously difficult challenge of manipulating deformable objects — work that has collectively garnered over 80 citations. Lian has also made meaningful advances in reinforcement learning for contact-rich manipulation, proposing methods to automatically learn dense reward functions that reduce the burden of manual reward engineering, an early version of which appeared in 2020 and a refined iteration in 2021. His benchmarking study of off-the-shelf robotic assembly solutions provided the community with valuable baselines for evaluating real-world manipulation performance. Further contributions include primitive-based skill learning from demonstrations, symbolic state estimation for contact-rich tasks, and zero-shot policy transfer via disentangled meta-reinforcement learning. Across his body of work, Lian consistently addresses the gap between theoretical robot learning and practical deployment, making his research particularly relevant to engineers and scientists working on industrial automation and intelligent robotic systems.
Research Focus
Key Achievements
Top Papers
- 1Robotic Cable Routing with Spatial Representation46 citations · 2022
- 2Cable Routing and Assembly using Tactile-driven Motion Primitives36 citations · 2023
- 3Learning Dense Rewards for Contact-Rich Manipulation Tasks24 citations · 2021
- 4Benchmarking Off-The-Shelf Solutions to Robotic Assembly Tasks23 citations · 2021
- 5
- 6
- 7
- 8Learning Dense Rewards for Contact-Rich Manipulation Tasks8 citations · 2020
- 9
- 10