Purdue ACRE Cornfield Dataset

This dataset, collected with Purdue-AgBot (P-AgBot) at the Agronomy Center for Research and Education (ACRE) during Summer 2023, supports research in autonomous navigation and crop monitoring in cornfields.

Overview

• Sensor Data: Includes 3D LiDAR, IMU, wheel encoders, and RTK GPS.
• Environments: ACRE cornfields under various weather conditions and growth stages. P-AgBot drove under the canopies in the cornfields.
• Challenges: Collected data in environments which contain hanging leaves and rough terrain.

Robot Platform and Sensors

• Unmanned Ground Vehicle (UGV): Clearpath Jackal (base_link)
• 3D LiDAR: Two Velodyne VLP-16 units mounted for horizontal (velodyne1) and vertical (velodyne2) scanning.
• IMU: Internal IMU in UGV (imu_link)
• Wheel Encoder: Internal wheel encoder in UGV
• RTK GPS: Emlid M2 (gps_link)

Accessing the Dataset

• Download the dataset here.

Data Description

Data is sorted by collection date and GPS availability, containing Rosbag (*.bag) files. Each file includes a comprehensive set of sensor measurements.

Figures

• Coordinate Frames of P-AgBot

• TF Tree Visualization

For details on sensor transformations, see static_transform.txt

Data Folders

Folder	Number of Files	Size (GB)
with_GPS	18	86.3
without_GPS	14	51.9

ROS Topics

Topic	Description	ROS Message Type
/cmd_vel	Robot linear/angular velocity	geometry_msgs/Twist
/gps/fix	RTK GPS measurements	sensor_msgs/NavSatFix
/imu/data	Robot IMU data	sensor_msgs/Imu
/ns1/velodyne_points	Point cloud from horizontal LiDAR velodyne1	sensor_msgs/PointCloud2
/ns2/velodyne_points	Point cloud from vertical LiDAR velodyne2	sensor_msgs/PointCloud2
/odometry/filtered	Filtered odometry from wheel encoders and IMU fusion	nav_msgs/Odometry
/tf	Sensor coordinate frames relationship	tf2_msgs/TFMessage

---

Citations

If you use this dataset for your research, please consider citing our works:

@ARTICLE{10494876,
  author={Kim, Kitae and Deb, Aarya and Cappelleri, David J.},
  journal={IEEE Robotics and Automation Letters}, 
  title={P-AgSLAM: In-Row and Under-Canopy SLAM for Agricultural Monitoring in Cornfields}, 
  year={2024},
  volume={9},
  number={6},
  pages={4982-4989},
  keywords={Feature extraction;Laser radar;Robots;Simultaneous localization and mapping;Three-dimensional displays;Point cloud compression;Monitoring;Agricultural automation;robotics and automation in agriculture and forestry;SLAM},
  doi={10.1109/LRA.2024.3386466}}

@INPROCEEDINGS{10341516,
  author={Deb, Aarya and Kim, Kitae and Cappelleri, David J.},
  booktitle={2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)}, 
  title={Deep Learning-Based Leaf Detection for Robotic Physical Sampling with P-AgBot}, 
  year={2023},
  volume={},
  number={},
  pages={8291-8297},
  keywords={YOLO;Deep learning;Three-dimensional displays;Robot kinematics;Robot vision systems;Crops;Grasping},
  doi={10.1109/IROS55552.2023.10341516}}

@ARTICLE{9810180,
  author={Kim, Kitae and Deb, Aarya and Cappelleri, David J.},
  journal={IEEE Robotics and Automation Letters}, 
  title={P-AgBot: In-Row & Under-Canopy Agricultural Robot for Monitoring and Physical Sampling}, 
  year={2022},
  volume={7},
  number={3},
  pages={7942-7949},
  keywords={Crops;Robots;Laser radar;Navigation;Monitoring;Three-dimensional displays;Autonomous robots;Field robots;agricultural automation},
  doi={10.1109/LRA.2022.3187275}}