About

Kaiyu Hang is a robotics researcher whose work sits at the intersection of robotic grasping, dexterous manipulation, and autonomous planning. With a research trajectory spanning over a decade, Hang has made foundational contributions to how robots perceive, grasp, and reposition objects in complex, uncertain environments. His early work on hierarchical fingertip spaces and dexterous grasping under shape uncertainty — now boasting over 100 citations — established principled frameworks for precision grasp synthesis when object geometry is incompletely known. Building on this foundation, Hang tackled increasingly sophisticated manipulation challenges: enabling robots to slide thin objects into graspable positions using compliant hands, perform finger gaiting to extend in-hand dexterity beyond kinematic constraints, and self-identify hand-object parameters for more controllable manipulation. Beyond grasping, Hang has advanced nonprehensile manipulation — pushing and rearranging objects without firmly holding them — using deep reinforcement learning, simulation-to-reality transfer, and Monte Carlo Tree Search. His benchmarking contributions in both grasp planning and in-hand manipulation have provided the community with standardized evaluation tools that are widely adopted. A recurring theme across his portfolio is making robotic manipulation robust, generalizable, and measurable — qualities that continue to shape the field's research agenda.

Research Focus

Key Achievements

16
H-Index
33
Papers
822
Total Citations
25
Avg Citations/Paper
🏆 Most Cited Paper
Dexterous grasping under shape uncertainty
106 citations · 2015
📈 Most Prolific Year: 2019 (12 Papers)
🤝 Key Collaborators: 61
🏛 Institutions: KTH Royal Institute of Technology, Rice University, Yale University, Hong Kong University of Science and Technology, Mater Health Services

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 14 days ago