CLARU
lerobotapache-2.0

OpenArm Hardware Modifications for Cloth Folding

Custom 3D-printable parts for modifying the OpenArm robot platform to enable bimanual cloth folding tasks, achieving 90% success rate on t-shirt folding.

Downloads82
Likes3

Why This Matters for Physical AI

These hardware modifications enable a bimanual robot to perform complex deformable object manipulation (cloth folding), which is a challenging physical AI task requiring dexterous control and tactile sensing.

Technical Profile

Modalities
tactile
Robot Embodiments
OpenArmbimanual
Environment
lab
Task Types
manipulationcloth_folding
License
apache-2.0
Part of the OpenArm Hardware Modifications for Cloth Folding family

Access

View on HuggingFace

Need custom tactile data?

Claru builds purpose-built datasets for lab applications with dense human annotations and quality assurance.

Request a Sample Pack

Related Datasets

FastUMI Pro – Multimodal Sample Dataset

Small-scale demonstration data from the FastUMI Pro multimodal sensing system, containing dozens of sample trajectories with synchronized RGB images, point clouds, pose tracking, and haptic feedback. Full-scale datasets are available upon request.

rgbpoint_cloudproprioceptiontactile
8K downloadsother

FuSe

FuSe contains 26,866 trajectories collected on a WidowX robot with visual, tactile, sound, and action data across multiple environments, annotated with natural language instructions.

rgbtactileaudioproprioception
2K downloads2025mit

FuSe

A multimodal robotics dataset containing 26,866 trajectories collected on a WidowX robot with visual, tactile, sound, and action data annotated with natural language instructions.

rgbtactileaudioproprioception
421 downloads2025MIT

VitaSet: Vision-Tactile VQA Dataset

A vision-tactile Visual Question Answering dataset for physical property reasoning combining RGB vision and tactile sensing, containing 5,145 human-verified QA pairs across hardness classification, material property description, and surface roughness classification tasks.

rgbtactile
295 downloads2025MIT

Crab Manipulation Dataset — Reward Labeled

A collection of 34 teleoperated robot manipulation demonstrations augmented with per-episode reward parameters for Sensitivity-Aware Reward-Weighted Flow Matching training. Includes clean demonstrations across egg placement, jar handling, and waffle manipulation tasks, plus negative examples with damage flags.

rgbtactileproprioception
292 downloadsapache-2.0

Crab Manipulation Dataset

Teleoperated demonstration dataset for contact-rich manipulation on the Crab bimanual mobile manipulator with dual SO-101 arms, collected at 15 Hz with 3 RGB cameras and dual 10×10 tactile force sensor matrices.

rgbproprioceptiontactile
252 downloadsapache-2.0