README.md

---
license: cc0-1.0
task_categories:
- tabular-classification
tags:
- robotics
- terrain-classification
- field-robotics
pretty_name: BorealTC
---

# Dataset Card for BorealTC

<!-- Provide a quick summary of the dataset. -->

This dataset contains IMU and wheel timeseries of data recorded with a Husky A200 UGV. It was recorded on five different types of terrains: snow, ice, silty loam, asphalt, flooring.

## Dataset Details

### Dataset Description

<!-- Provide a longer summary of what this dataset is. -->

Recorded with a Husky A200 wheeled UGV, `BorealTC` contains 116 min of Inertial Measurement Unit (IMU), motor current, and wheel odometry data, focusing on typical boreal forest terrains, notably snow, ice, and silty loam. The dataset also includes experiments on asphalt and flooring. All runs were recorded in Forêt Montmorency and on the main campus of Université Laval, Quebec City, Quebec, Canada.


- **Curated by:** Northern Robotics Laboratory, Université Laval, Québec, Canada
- **License:** [CC0 1.0 Universal](https://creativecommons.org/publicdomain/zero/1.0)

### Dataset Sources

<!-- Provide the basic links for the dataset. -->

- **Repository:** [norlab-ulaval/BorealTC](https://github.com/norlab-ulaval/BorealTC)
- **Paper:** [10.1109/IROS58592.2024.10801407](https://doi.org/10.1109/IROS58592.2024.10801407)
- **Page:** [BorealTC](https://norlab-ulaval.github.io/BorealTC)

## Uses

<!-- Address questions around how the dataset is intended to be used. -->

This data was intended for terrain classification problems.

### Direct Use

<!-- This section describes suitable use cases for the dataset. -->

This dataset could be used as example data for sensor timeseries processing.

## Dataset Structure

<!-- This section provides a description of the dataset fields, and additional information about the dataset structure such as criteria used to create the splits, relationships between data points, etc. -->

Each folder contains data for runs recorded on a specific terrain class. The data for each run is included in two CSV files: `imu_*.csv` and `pro_*.csv`:

```sh
borealtc
├── CLASS1
│   ├── imu_00.csv
│   ├── imu_01.csv
│   ├── ...
│   ├── pro_00.csv
│   ├── pro_01.csv
│   └── ...
└── CLASS2
    ├── imu_00.csv
    ├── imu_01.csv
    ├── ...
    ├── pro_00.csv
    ├── pro_01.csv
    └── ...
```

Each `imu` file contains IMU-recorded rotational velocities and linear accelerations.

```csv
time,wx,wy,wz,ax,ay,az
0.0,0.0015195721884953,0.0040130227245162,-0.0070785057037968,1.4258426214785636,-0.0832771308374386,9.609228803438713
...
```

Each `pro` file contains motor currents and wheel velocities recorded by the wheel service of the Husky.

```csv
time,curL,curR,velL,velR
0.0,1.78,2.57,0.0236220472440944,0.0236220472440944
...
```


## Dataset Creation

### Curation Rationale

<!-- Motivation for the creation of this dataset. -->

This dataset aims at collecting terrain data on terrains typical of boreal forests.



#### Data Collection and Processing

<!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->

This dataset was recorded with a Husky A200 wheeled UGV on five terrains.

## Citation

<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->

**BibTeX:**

```bibtex
@inproceedings{LaRocque2024,
  title     = {{Proprioception Is All You Need: Terrain Classification for Boreal Forests}},
  url       = {http://dx.doi.org/10.1109/IROS58592.2024.10801407},
  doi       = {10.1109/iros58592.2024.10801407},
  booktitle = {2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  publisher = {IEEE},
  author    = {LaRocque,  Damien and Guimont-Martin,  William and Duclos,  David-Alexandre and Giguère,  Philippe and Pomerleau,  Fran\c{c}ois},
  year      = {2024},
  month     = oct,
  pages     = {11686–11693}
}
```

## Contributions

Thanks to @WillGuimont and @Asers387 for the help in curating this dataset.