FIGAROH Examples

Examples for the FIGAROH PLUS library (robot dynamics identification and geometric calibration).

Install

pip install figaroh
pip install -r requirements.txt

If you use conda, some dependencies may be easier to install via conda:

conda install -c conda-forge pinocchio cyipopt

Run

Most scripts assume you run them from inside the corresponding robot folder:

cd examples/ur10
python calibration.py

Basic workflow

1. Choose an example under examples/<robot>/.
2. Review the YAML files under examples/<robot>/config/.
3. Place or update CSV logs under examples/<robot>/data/ (or update paths in the YAML).
4. Run one of: calibration.py, identification.py, optimal_config.py, optimal_trajectory.py (if present).
5. Review printed results/plots. If applicable, use update_model.py to materialize estimated parameters.
6. Add --html-report (calibration/identification scripts) for a shareable HTML diagnostic report, and --verify (identification scripts) for a machine-readable pass/fail verdict you can gate CI on. See each robot's README for the exact flags it supports, and FIGAROH's Reporting & Verification guide for the full walkthrough (HTML reports, verify(), and comparing two runs offline).
7. Add --wls/--no-wls (identification scripts) to override the config's identification.problem.wls value and refine the OLS base-parameter estimate with weighted least squares before quality metrics are computed (Staubli TX40's config defaults this on; UR10/TIAGo default off).
8. By default (unless run with --no-archive), each run is archived to a timestamped results/runs/<asset>/<task>/<timestamp>/ directory containing a config snapshot, a provenance record (git commit, config hash, timestamps), and the HTML report if generated — instead of overwriting a single results/ path — and appends a summary line to results/runs/index.jsonl.

Data format

Examples use CSV logs for measurements and trajectories. The required files/columns depend on the robot and task; see each example README for the expected inputs.

Examples

• UR10 (manipulator): examples/ur10/README.md
• TIAGo (mobile manipulator): examples/tiago/README.md
• TALOS (humanoid, torso/arm chain): examples/talos/README.md
• Staubli TX40 (manipulator): examples/staubli_tx40/README.md
• Templates and config starting points: examples/templates/README.md

Common layout (per robot)

Most robot folders follow this pattern:

{robot}/
  calibration.py            # kinematic calibration (if present)
  identification.py         # dynamic identification (if present)
  optimal_config.py         # optimal measurement configurations (if present)
  optimal_trajectory.py     # exciting trajectories for identification (if present)
  config/                   # YAML configuration files
  data/                     # CSV logs / measurement data
  urdf/                     # robot URDF(s) used by the scripts
  utils/                    # robot-specific helper classes

Creating a new example

Use the scaffold script to create a new robot folder based on the TIAGo template:

cd examples
./create_example.sh <robot_name>

The generated scripts are placeholders that point you back to the TIAGo example for a complete reference implementation.

Citation

If you use these examples in your research, please cite the main FIGAROH paper:

@inproceedings{nguyen2023figaroh,
  title={FIGAROH: a Python toolbox for dynamic identification and geometric calibration of robots and humans},
  author={Nguyen, Dinh Vinh Thanh and Bonnet, Vincent and Maxime, Sabbah and Gautier, Maxime and Fernbach, Pierre and others},
  booktitle={IEEE-RAS International Conference on Humanoid Robots},
  pages={1--8},
  year={2023},
  address={Austin, TX, United States},
  doi={10.1109/Humanoids57100.2023.10375232},
  url={https://hal.science/hal-04234676v2}
}

License

Apache License 2.0. See LICENSE.

Support

• Open an issue in this repository for example-specific questions.
• Open an issue in the main FIGAROH repository: https://github.com/thanhndv212/figaroh-plus/issues