Faless2026apache-2.0
Harvest Apples with AGILex Piper Sim DAGGER
A robotics dataset of apple-picking tasks collected using the Piper robot platform with sim-to-real transfer via DAGGER (Dataset Aggregation). The dataset contains 706 episodes of manipulation trajectories recorded at 25 fps with multi-view RGB observations and joint state information.
Downloads31
Episodes706
Hours13.1181
Why This Matters for Physical AI
This dataset demonstrates sim-to-real transfer for robotic manipulation through DAGGER imitation learning, providing a benchmark for training policies on complex fruit-picking tasks with multi-view visual feedback.
Technical Profile
- Modalities
- rgbproprioception
- Robot Embodiments
- Piper
- Action Space
- 7-DoF: joint_1.pos, joint_2.pos, joint_3.pos, joint_4.pos, joint_5.pos, joint_6.pos, gripper.pos
- Environment
- simulation
- Task Types
- apple-pickingmanipulation
- Episodes
- 706
- Total Hours
- 13.1181
- Data Format
- parquet
- License
- apache-2.0
Access
Need custom rgb data?
Claru builds purpose-built datasets for simulation applications with dense human annotations and quality assurance.
Request a Sample Pack