Teleoperation Tabletop Dataset
Robot teleoperation data for tabletop manipulation — sorting, stacking, tool use — with synchronized camera-action-force triplets for training general-purpose policies.
Dataset at a Glance
Comparison with Public Datasets
How Claru's dataset compares to publicly available alternatives.
| Dataset | Clips | Hours | Modalities | Environments | Annotations |
|---|---|---|---|---|---|
| RoboTurk | 111K | ~200 | RGB | Lab | Actions |
| Bridge Data V2 | 60K | ~200 | RGB | Lab counter | Actions, language |
| Claru Tabletop | 80K+ | 500+ | RGB, Depth, F/T | 25+ task families | Actions, forces, contacts, language, success |
Use Cases
Foundation Manipulation Models
General-purpose tabletop policies that transfer across objects and tasks. Example models: Octo, RT-X, OpenVLA.
Dexterous Manipulation
Fine-grained finger control for precise manipulation and assembly. Example models: DexNet, DexMV, UniDexGrasp.
Language-Conditioned Policies
Following natural language instructions for diverse tabletop tasks. Example models: RT-2, SayCan, CLIPort.
Key References
How Claru Delivers This Data
Claru's tabletop dataset covers 25+ task families with data from multiple robot embodiments, enabling cross-embodiment transfer learning. Force/torque sensing captures contact dynamics critical for precision manipulation.
Frequently Asked Questions
25+ families: sorting (color, shape, size), stacking, tool use (spatula, tongs, scissors), pouring, insertion, drawer opening, button pressing, and multi-step assembly.
Each family uses 50-200 unique objects varying in shape, size, material, and color.
A subset of 15K+ trajectories includes bimanual manipulation using dual-arm setups.
Request a Sample Pack
Get a curated sample of teleoperation tabletop data with full annotations to evaluate for your project.