Papers

13

Total Citations

197

H-Index

8

About

Jens Lundell is a roboticist whose research lies at the intersection of dexterous manipulation, grasp synthesis, and scene understanding. His work is distinguished by a focus on enabling robots to handle complex, real-world scenarios—from cluttered bins to deformable objects—by integrating learning-based methods with explicit reasoning about uncertainty and physics. Lundell’s most influential contribution, “Towards Robust 6-DoF Multi-Finger Grasping in Clutter with Explicit Scene Understanding” (65 citations), set a new standard for robust grasping by combining deep learning with structured scene analysis. He has also pioneered the use of diffusion models for dexterous grasp generation (DexDiffuser, 24 citations) and advanced the field of active perception through visuo-haptic object shape completion (25 citations). His work on deformation-aware grasping and imitation learning for assembly tasks demonstrates a rare ability to bridge data-driven methods with principled robotics. With over 180 total citations and a portfolio spanning tactile sensing, POMDP planning, and navigation under uncertainty, Lundell is shaping the next generation of autonomous manipulation systems that are both intelligent and physically capable.

Research Focus

Key Achievements

8
H-Index
13
Papers
197
Total Citations
15
Avg Citations/Paper
🏆 Most Cited Paper
Towards Robust 6-DoF Multi-Finger Grasping in Clutter with Explicit Scene Understanding
65 citations · 2022
📈 Most Prolific Year: 2022 (4 Papers)
🤝 Key Collaborators: 25
🏛 Institutions: Aalto University, KTH Royal Institute of Technology, University of Technology

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 · 13 days ago