Contact onset and grip evolution are under-captured
Most datasets miss touch timing, pressure trends, and subtle adjustments that matter in dexterous manipulation.
Touch-aware demonstration data for the next generation of dexterous robots.
Robots have five senses. We are building the sixth.
Most datasets miss touch timing, pressure trends, and subtle adjustments that matter in dexterous manipulation.
Standalone sensors and recording scripts often fail calibration, synchronization, and dataset reliability requirements.
Robot learning teams want high-value datasets, not multi-month setup and QC work for each collection effort.
For contact-rich tasks, the next performance gains come from better multimodal demonstrations, not model scale alone.
Hardware, sync, calibration, packaging - one stack, four stages. Robot-learning teams get aligned episodes, not raw folders.
High-rate contact and pressure-aligned streams with per-channel calibration — where touch matters for the task.
First-person RGB aligned to what the demonstrator sees — stable exposure for long household runs.
Per-frame depth aligned to ego timebase for geometry, reach, and clutter around the hands.
Articulated hand state and grasp phases for contact-rich manipulation — not just 2D boxes in frame.
Linear acceleration, angular rates, and movement cues that characterize inertia, rhythm, and effort during the task.
Secondary viewpoints for occlusion recovery, tool use, and context beyond the ego cone (roadmap / program-dependent).
Task and subtask boundaries, contact phases, QC flags — plus optional timestamped, frame-aligned text narration paired to video for richer training supervision.
Binary or graded success, failure modes, and segment-level tags for imitation and evaluation.
We document where each signal is reliable, how drift is handled, and what should never be treated as ground-truth force.
Pressure proxies, contact timing, and failure flags are defined so policy teams know exactly what each dimension means.
Episodes land in the formats your trainers expect, with QC metrics and assumptions — not a dump of raw sensor folders.