Jennifer Grannen

About

Jennifer Grannen is a leading researcher in robotic manipulation of deformable objects, with a focus on the challenging domains of ropes, cables, and fabrics. Her work addresses the fundamental difficulty of controlling objects with infinite-dimensional configuration spaces, complex dynamics, and self-occlusion. Grannen’s major contributions include developing learning-based methods that use dense object descriptors trained on synthetic data to establish visual correspondences, enabling robots to smooth, fold, and untangle real fabrics and ropes. Her highly cited paper "Learning Rope Manipulation Policies Using Dense Object Descriptors Trained on Synthetic Depth Data" (103 citations) demonstrates how simulated training can transfer to real-world manipulation. She has also advanced the field of knot untangling with algorithms like IRON-MAN for disentangling multiple cables. More recently, Grannen has expanded into assistive robotics, with notable work on in-mouth robotic bite transfer using visual and haptic sensing, and bimanual scooping policies for food acquisition. Her research consistently bridges simulation and reality, achieving robust performance on complex, real-world tasks.

Research Focus

Key Achievements

Top Papers

1. 1
   Learning Rope Manipulation Policies Using Dense Object Descriptors Trained on Synthetic Depth Data

   103 citations · 2020
2. 2
   Learning Dense Visual Correspondences in Simulation to Smooth and Fold Real Fabrics

   55 citations · 2021
3. 3
   Untangling Dense Non-Planar Knots by Learning Manipulation Features and Recovery Policies

   26 citations · 2021
4. 4
   Disentangling Dense Multi-Cable Knots

   19 citations · 2021
5. 5
   Learning to Smooth and Fold Real Fabric Using Dense Object Descriptors Trained on Synthetic Color Images

   19 citations · 2020
6. 6
   Learning Rope Manipulation Policies Using Dense Object Descriptors Trained on Synthetic Depth Data

   16 citations · 2020
7. 7
   Learning Dense Visual Correspondences in Simulation to Smooth and Fold Real Fabrics

   16 citations · 2020
8. 8
   In-Mouth Robotic Bite Transfer with Visual and Haptic Sensing

   13 citations · 2023
9. 9
   Learning Bimanual Scooping Policies for Food Acquisition

   9 citations · 2022
10. 10
    Untangling Dense Knots by Learning Task-Relevant Keypoints

    5 citations · 2020

Key Collaborators

Contact & Links