Trajectory Optimization for In-Hand Manipulation with Tactile Force Control [IROS2025]

Official Code Repository for Trajectory Optimization for In-Hand Manipulation with Tactile Force Control.
Minimal, uv-managed Python project for in-hand rolling trajectory optimization (CasADi + IPOPT) with a MeshCat visualizer.

Haegu Lee¹, Yitaek Kim¹, Victor Melbye Staven¹, Christoffer Sloth¹

¹University of Southern Denmark

This repo contains:

• finger.py – planar finger + circular object model (CasADi)
• main.py – builds/solves the NLP and saves finger_solution.npz
• animate.py – MeshCat animation of the saved trajectory
• README.md – this file

---

Requirements

• Python 3.10 – 3.12
• uv (dependency manager)
• A CasADi build with IPOPT (standard wheels typically include it)

---

Install with uv

# 1) Create a virtual environment (managed by uv)
uv venv --python 3.12

# 2) Install dependencies from pyproject.toml (creates/updates lockfile)
uv lock
uv sync

How to run

Run the top-level scripts directly (no console entry points, by design):

# Optimize and save results to finger_solution.npz
uv run python main.py

# Visualize the saved result in MeshCat
uv run python animate.py

The result should look like this.

Repository Structure

rolling_trajectory_optimization
├── animate.py
├── editorconfig
├── finger.py
├── finger_rolling.gif
├── finger_solution.npz
├── gitignore
├── main.py
├── pyproject.toml
├── README.md
└── uv.lock

Citation

If you find our codes useful in your work, please cite our paper:

@article{lee2025trajectory,
  title={Trajectory Optimization for In-Hand Manipulation with Tactile Force Control},
  author={Lee, Haegu and Kim, Yitaek and Staven, Victor Melbye and Sloth, Christoffer},
  journal={arXiv preprint arXiv:2503.08222},
  year={2025}
}

Contact

If you have any questions, feel free to contact me through email (haeg@mmmi.sdu.dk)!