{"product_id":"yumi","title":"Yumi","description":"\u003cp\u003e\u003cmeta charset=\"utf-8\"\u003e\u003cspan style=\"text-decoration: underline;\"\u003e\u003cspan class=\"s1\" style=\"color: rgb(255, 42, 0); text-decoration: underline;\"\u003e\u003cb\u003eOrder placed now will ship in 15 business days.\u003c\/b\u003e\u003c\/span\u003e\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003eYumi is a handheld gripper built around the same Linear Gripper mechanism used on i2RT's YAM robot arms. Every demonstration you capture transfers seamlessly to YAM robots for policy training — no retargeting, no data conversion gap.\u003cbr\u003ePair Yumi with HTC Vive SteamVR Trackers or Meta Quest Touch Controllers for 6DoF pose tracking, capture synchronized visual streams and gripper joint states, and export ready-to-use datasets for imitation learning. Single-hand and bimanual capture are both supported.\u003c\/p\u003e\n\u003cp\u003e\u003cstrong\u003eHow It Works\u003c\/strong\u003e\u003cbr\u003e1. Capture — Hold Yumi and perform your manipulation tasks. The high-speed global shutter camera records the scene while your tracking hardware outputs 6DoF pose and gripper joint feedback in real time.\u003cbr\u003e2. Export — Use the Yumi SDK to export structured, ready-to-use datasets.\u003cbr\u003e3. Train — Load demonstrations directly into policy training for your YAM robotic arms, with native compatibility across LeRobot, ALOHA, and Isaac Sim.\u003c\/p\u003e\n\u003cp\u003e\u003cstrong\u003eKey Features\u003c\/strong\u003e\u003cbr\u003e- Same Linear Gripper hardware as the i2RT YAM series — captured data fits YAM deployment perfectly\u003cbr\u003e- High-speed global shutter camera, up to 400 fps for fast, dynamic motions\u003cbr\u003e- Lightweight two-finger parallel gripper at an estimated 500 g\u003cbr\u003e- Dual tracking options: HTC Vive SteamVR (sub-millimeter static precision) or Meta Quest inside-out tracking\u003cbr\u003e- Gripper joint state feedback up to 500 Hz\u003cbr\u003e- URDF export supported\u003cbr\u003e- USB-C wired design — no onboard battery to manage or charge\u003cbr\u003e- Bimanual human demonstration collection supported\u003c\/p\u003e\n\u003cp\u003e\u003cstrong\u003eWho It's For\u003c\/strong\u003e\u003cbr\u003eRobotics research labs, embodied AI R\u0026amp;D teams, robot learning engineers, and dataset providers.\u003c\/p\u003e\n\u003cp\u003e\u003cstrong\u003eApplications\u003c\/strong\u003e\u003cbr\u003e- In-the-wild human demonstration collection\u003cbr\u003e- Behavior cloning and imitation learning for robotic manipulation\u003cbr\u003e- Bimanual skill training for YAM robots\u003cbr\u003e- Rapid policy prototyping\u003c\/p\u003e\n\u003cp\u003e\u003cstrong\u003eSpecifications\u003c\/strong\u003e\u003c\/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003eCamera\u003c\/strong\u003e\u003cbr\u003e- Sensor: 1\/2.0\" global shutter, 4.8µm × 4.8µm pixel size\u003cbr\u003e- Max resolution: 1280 × 1024 @ 200 fps (MJPEG) \/ 100 fps (YUV)\u003cbr\u003e- High-speed mode: 640 × 480 \/ 640 × 512 @ 400 fps\u003cbr\u003e- Output formats: MJPEG, YUV2 (YUYV)\u003cbr\u003e- Lens: f = 3.25mm, F2.7 aperture, 154.2° diagonal FOV, M12×P0.5 mount\u003cbr\u003e- SNR: 37.5 dB\u003cbr\u003e- Dynamic range: 77 dB\u003cbr\u003e- Minimum illumination: 0.5 lux\u003cbr\u003e- Auto features: AEC, AWB, AGC supported\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eGripper\u003c\/strong\u003e\u003cbr\u003e- Type: Two-finger parallel jaw gripper (same Linear Gripper as i2RT YAM)\u003cbr\u003e- Estimated max gripping force: 2 kg (preliminary estimate, pending formal testing)\u003cbr\u003e- Max opening: 95 mm\u003cbr\u003e- Min opening: 0 mm (fine enough to grip thin paper)\u003cbr\u003e- Joint rotation precision: ±0.1°\u003cbr\u003e- Max output feedback frequency: 500 Hz\u003cbr\u003e- Estimated weight: 500 g\u003cbr\u003e- Dimensions: 186 × 158 × 207 mm\u003cbr\u003e- Sensing: No tactile or force sensing\u003cbr\u003e- Power: 5V, ~3W (camera and HTC tracker included)\u003cbr\u003e- Communication interface: USB-C, wired connection, no onboard battery\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003ePose Tracking Module\u003c\/strong\u003e\u003cbr\u003e- Compatible hardware: HTC Vive SteamVR Trackers \/ Meta Quest Touch Controllers\u003cbr\u003e- Tracking degrees of freedom: 6DOF\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eHTC Vive SteamVR 2.0 Tracker (base station required):\u003c\/strong\u003e\u003cbr\u003eSub-millimeter positioning accuracy (static) · 90 Hz pose update rate · Recommended tracking volume up to 10m × 10m room scale\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eMeta Quest Touch \/ Touch Pro Controllers (inside-out):\u003c\/strong\u003e\u003cbr\u003e~±2–5mm positioning accuracy (static, varies with occlusion and lighting) · 90 Hz pose update rate, 240 Hz on-controller IMU sampling · Tracking volume limited by the Quest headset's camera field of view\u003cbr\u003e- Onboard IMU: None — IMU data is sourced from the HTC Vive \/ Meta Quest controller hardware\u003cbr\u003e- Calibration: Follows the native HTC Vive \/ Meta Quest calibration workflow\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eSoftware \u0026amp; Ecosystem\u003c\/strong\u003e\u003cbr\u003e- SDK: Available\u003cbr\u003e- URDF export: Supported\u003cbr\u003e- Supported frameworks: LeRobot, ALOHA, Isaac Sim\u003cbr\u003e- Bimanual data capture: Supported\u003cbr\u003e- Data pipeline: Real-time preview and data validation can be implemented via secondary development using the open-source SDK\u003cbr\u003e- Multi-sensor time synchronization accuracy: TBD\u003cbr\u003e- Operating temperature: Indoor ambient temperature (human-operated device, no formal temperature rating)\u003c\/li\u003e\n\u003c\/ul\u003e\n\u003cp\u003e\u003cstrong\u003eWhat is in the box?\u003c\/strong\u003e\u003cbr\u003e- i2RT Yumi Handheld Gripper Assembly\u003cbr\u003e- Onboard Global Shutter Camera\u003cbr\u003e- USB-C Data \u0026amp; Power Cable\u003cbr\u003e- SDK \u0026amp; Documentation download access\u003cbr\u003e\u003cspan style=\"text-decoration: underline;\"\u003e\u003cem\u003e\u003cstrong\u003e\u003c\/strong\u003e\u003c\/em\u003e\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan style=\"text-decoration: underline;\"\u003e\u003cem\u003e\u003cstrong\u003eNote: HTC Vive \/ Meta Quest controllers and trackers are not included and must be supplied by the user.\u003c\/strong\u003e\u003c\/em\u003e\u003c\/span\u003e\u003c\/p\u003e\n\u003cp\u003e\u003cspan style=\"text-decoration: underline;\"\u003e\u003cem\u003e\u003cstrong\u003e \u003c\/strong\u003e\u003c\/em\u003e\u003c\/span\u003e\u003cem\u003eIf you have any questions about warranty-related matters, please refer to our \u003ca href=\"https:\/\/i2rt.com\/pages\/warranty-policy\"\u003eWarranty Policy\u003c\/a\u003e\u003cstrong\u003e \u003c\/strong\u003efor more details.\u003c\/em\u003e\u003c\/p\u003e","brand":"I2RT Robotics","offers":[{"title":"A set","offer_id":50380362285313,"sku":null,"price":599.0,"currency_code":"USD","in_stock":true},{"title":"Left","offer_id":50380362219777,"sku":null,"price":369.0,"currency_code":"USD","in_stock":false},{"title":"Right","offer_id":50380362252545,"sku":null,"price":369.0,"currency_code":"USD","in_stock":false}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0733\/1276\/4161\/files\/04_0fc6b3a0-d494-4588-af1e-e04e41466f87.png?v=1790152126","url":"https:\/\/i2rt.com\/products\/yumi","provider":"I2RT Robotics","version":"1.0","type":"link"}