# WMU AD Assurance Lab Drives 2019-2025: Multi-Camera, Automotive Radar, CAN, GNSS and Road-Weather Recordings (Western Michigan University)

## Facts

- Provider: AD Assurance Lab, Western Michigan University
- License: CC-BY-4.0 (https://huggingface.co/datasets/AD-Assurance-Lab/lab-drives)
- Size: 473 files
- Format: Parquet, MP4, JSON (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/AD-Assurance-Lab/lab-drives
- Access: public, readable without an account
- Category: Robotics
- Source: https://huggingface.co/datasets/AD-Assurance-Lab/lab-drives
- Page: https://scigantic.com/archive/47980c8c-68d1-5023-bfeb-31411b2fd82b
- Keywords: Robotics, LiDAR, Parquet, MP4, JSON, Hugging Face

## About

Smaller automated-driving recordings from the AD Assurance Lab at Western Michigan University, companion to the lab's Ouster LiDAR drives. Collections: a 27-minute December 2023 urban drive in Kalamazoo with six Basler cameras (1440x1080 at about 15 fps), GNSS, road-weather sensing and vehicle CAN bus; eight 2023-24 radar test bags with Aptiv ESR radar, one camera and CAN; radar-only highway and city drives from 2020-21; two 2022 campus drives with the ZED 2i stereo camera's left eye; a 2023 Basler camera test; two camera-LiDAR chessboard calibration recordings; a 2025 campus shuttle drive; 13 static 2D-lidar runs in four weather conditions from 2019; and phone videos and photos of a snowy 2020 highway drive. Sensor tables on the card give make, model, rates and the source paper for each collection. Data are Parquet tables, JSON metadata, MP4/MKV video and JPEG images in one folder per collection, about 33 GB in 473 files (measured 2026-11-21); read with pyarrow or pandas. Use it for camera and radar perception, sensor calibration and adverse-weather studies. Released under CC BY 4.0 by the lab; cite the paper given for the data you use.

## Load the data

```python
from datasets import load_dataset
ds = load_dataset("AD-Assurance-Lab/lab-drives")
print(ds)
```

## Data structure

File types in sampled files: Parquet, MP4, JSON

Example file: `basler_campus_2023-10/2023-10-16-17-43-58/camera.frames.parquet`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/47980c8c-68d1-5023-bfeb-31411b2fd82b
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/47980c8c-68d1-5023-bfeb-31411b2fd82b
- Over MCP: `get_schema_card` with id `47980c8c-68d1-5023-bfeb-31411b2fd82b` at https://api.scigantic.com/mcp

## Related topics

- [Parquet datasets](https://scigantic.com/topics/parquet)
- [JSON datasets](https://scigantic.com/topics/json)
- [LiDAR datasets](https://scigantic.com/topics/lidar)
- [Robotics datasets](https://scigantic.com/topics/robotics)
- [MP4 datasets](https://scigantic.com/topics/mp4)
