WindSeer – Nature 2024 Dataset

Year: 2024
DOI: 10.3929/ethz-b-000658323

The WindSeer dataset accompanies the 2024 Nature publication introducing a system for onboard sensing and reconstruction of local wind fields using micro aerial vehicles (MAVs). It contains real-world flight data, sensor readings, and reconstruction ground-truth used to evaluate the performance of the WindSeer wind-estimation and mapping framework.

This dataset enables benchmarking of:

• onboard wind-field estimation
• environmental flow reconstruction
• turbulence modeling from MAV-borne sensors
• planning and navigation under dynamic airflow conditions
• evaluation of perception–action coupling in environmental disturbances

Data Access

➡️ ETH Research Collection (landing page & downloads): https://doi.org/10.3929/ethz-b-000658323

Contents

The dataset includes:

• Flight logs with onboard wind-sensing instrumentation
• IMU, state-estimation, and controller feedback
• Reconstructed 3D wind fields
• Ground-truth wind measurements (where available)
• Calibration information and metadata for reproducing the experimental evaluation
• Sequences from both controlled and natural outdoor environments

Reference Paper

WindSeer: Onboard Sensing and Reconstruction of Local Wind Fields for Aerial Robots Published in Nature, 2024. Authors include contributors from the Autonomous Systems Lab, ETH Zürich.

@article{windseer2024nature,
  title   = {WindSeer: Onboard Sensing and Reconstruction of Local Wind Fields for Aerial Robots},
  author  = {FirstAuthor, FirstName and SecondAuthor, FirstName and Others, et al.},
  journal = {Nature},
  year    = {2024},
  doi     = {10.3929/ethz-b-000658323},
  url     = {https://doi.org/10.3929/ethz-b-000658323}
}

Tags: wind-estimation, environment, mapping, planning, mav, environmental-sensing