휴머노이드 AI 연구를 위한 풀스택 개발 플랫폼
H2 Plus는 NVIDIA Jetson Thor 기반의 고성능 온보드 컴퓨팅과 비전·촉각 센싱, 5지 덱스터러스 핸드를 통합해 데이터 수집부터 학습, 시뮬레이션, 정책 실행까지 지원합니다.
Captures high-quality robot demonstration data for training and policy development.
Supports humanoid reasoning, learning and multitask behavior.
Simulates, trains, tests and evaluates robot policies before real-world deployment.
Moves trained policies onto robots.
Runs real-time, on board inference and control.
| Mechanical Dimensions | Height × Width × Depth (Standing) | 1820 x 456 x 218mm |
| Weight with Battery | About 70kg | |
| Degrees of Freedom (Number of Joint Motors) | 31 | |
| Single Leg Degrees of Freedom | 6 | |
| Single Arm Degrees of Freedom | 7 | |
| Waist Degrees of Freedom | 3 | |
| Head Degrees of Freedom | 2 | |
| Joint Output Bearings | Industrial-grade crossed roller bearings (high precision, high load capacity) | |
| Joint Motors | Low-inertia high-speed inner-rotor permanent magnet synchronous motors (better response speed and heat dissipation) | |
| Maximum Arm Joint Peak Torque [1] | 120N.m | |
| Maximum Leg Joint Peak Torque [1] | 360N.m | |
| Maximum Arm Payload [1] | Peak: approx. 15kg; Rated: approx. 7kg | |
| Lower Leg + Upper Leg Length | 1045mm | |
| Forearm + Upper Arm Length | 690mm | |
| Main Body Materials | Aerospace-grade aluminum + titanium alloy + high-strength engineering plastics | |
| Electrical Characteristics | Cooling System | Air Cooling |
| Power Supply Method | Low-Resistance High-Quality Lithium Battery | |
| Battery Voltage | Maximum Voltage: 75.6V | |
| Base Computing Power | PC1: Intel Core i5 (Platform Functions) PC2: Intel Core i7 (User Development) |
|
| Perception Sensors | Wide-FOV Humanoid Binocular Camera | |
| External Interfaces | USB3.0 x 3 USB+DP x 3 GMSL x 4 Gigabit Ethernet x 4 USB2.0 x 2 RS485 x 4 CAN x 2 Power Outputs：12V x 1；24V x 2；BAT x 1 |
|
| Voice Interaction | Voice interaction via array microphones and high-power speakers | |
| WiFi 6 、Bluetooth 5.2 | YES | |
| Accessories | High-Performance Computing Module | NVIDIA Jetson™ T5000 |
| Smart Battery(Quick Release) | YES | |
| Charger | YES | |
| Handheld Remote Controller | YES | |
| Compute Module | AI Performance | 2070 TFLOPS (FP4—sparse) |
| GPU | 2560-core NVIDIA Blackwell Architecture GPU with 5th-Generation Tensor Cores Multi-Instance GPU with 10 TPCs |
|
| GPU Max Frequency | 1.57 GHz | |
| CPU | 14-core Arm® Neoverse®-V3AE 64-bit CPU 1 MB L2 cache per core 16 MB shared system L3 cache |
|
| CPU Max Frequency | 2.6 GHz | |
| Vision Accelerator | 1x PVA v3 | |
| Memory | 128 GB 256-bit LPDDR5X 273 GB/s |
|
| Storage | Supports NVMe through PCle Supports SSD through USB3.2 |
|
| Video Encode | 2x NVENC | |
| Video Decode | 2x NVDEC | |
| Vision Module | Head Binocular Camera | Horizontal FOV 140°±3; Vertical FOV 102°±3; Aperture 2.0; Focal Length 2.67mm; Depth of Field 0.4m-INF; global shutter; |
| Wrist Cameras (Optional) | 3MP ultra-wide-angle, HFOV 196°, VFOV 154°, Global shutter | |
| Sharpa Wave Dexterous Hand (Optional) | Dexterous Hand Weight | 1.3KG |
| Dexterous Hand Dimensions | 208 x 90 x 50mm | |
| Active Degrees of Freedom per Dexterous Hand | 22 | |
| Dexterous Hand Payload | 30Kg | |
| Dexterous Hand Grip Force | 150N | |
| Dexterous Hand Fingertip Force | 20N | |
| Dexterous Hand Operating Speed | >4Hz | |
| Fingertip Positional Repeatabillity | ±1mm | |
| Number of Dexterous Hand Tactile Pixels | >1000 tactile pixels per fingertip | |
| Dexterous Hand Force Detection Range | Range 0–30N, Sensitivity 0.005N | |
| Other | Battery Life | About 3h |
| Smart OTA Upgrade | YES | |
| Secondary Development [2] | YES | |
| Warranty Period [3] | 12 Months |
