SINGAPORE, June 1, 2026 /PRNewswire/ — Sharpa today announced that the NVIDIA Isaac GR00T Reference Humanoid Robot is the first dexterous humanoid robot reference design built on the Isaac GR00T development platform to feature Sharpa’s Wave robot hands for dexterous, tactile manipulation. This validated configuration gives developers and researchers a state-of-the-art humanoid robot for building, fine-tuning and deploying robot skills faster — reducing setup time from days to hours.
As demand for general-purpose humanoids accelerates, developers still face a fragmented workflow spanning hardware integration, data collection, simulation, training, evaluation and deployment.
The NVIDIA Isaac GR00T Reference Humanoid Robot unifies development by bringing a Unitree H2 Plus humanoid robot and Sharpa Wave tactile five-finger hands (the “body”), with NVIDIA Jetson Thor-powered onboard compute and Isaac GR00T software and workflows (the “brain”) into a single integrated reference design, helping research teams move faster from robot bring-up to skill development and real-world validation.
“Our vision is to make robots genuinely productive — by advancing fine manipulation skills through dexterous, tactile hardware and the AI models that power them”, said David Li, Founder of Sharpa. “Partnering with NVIDIA on a humanoid robot reference design and end-to-end development solution is a meaningful step toward deploying robots that can perform real work, in real settings.”
“Dexterous hands are essential for humanoid robots to perform useful manipulation tasks in the real world,” said Spencer Huang, Director of Product for Robotics at NVIDIA. “With Sharpa Wave tactile five-finger hands, NVIDIA Isaac GR00T and Isaac Teleop, the NVIDIA Isaac GR00T Reference Humanoid Robot gives developers a reference design for training robot skills that require touch, control and precision.”
A State-of-the-Art Dexterous Humanoid for Physical AI Development
The Sharpa Wave-enhanced NVIDIA Isaac GR00T Reference Humanoid Robot brings together the essential building blocks for frontier humanoid research into a single integrated system — pairing a human-scale robot body with dexterous manipulation, tactile sensing, whole-body control and onboard AI compute.
The reference design features:
- Dual Sharpa Wave tactile five-finger hands, with 22 degrees of freedom per hand, bringing the robot to 75 total degrees of freedom across hands and body and enabling dexterous manipulation at human scale.
- A high-resolution Digital Tactile Array with >1,000 pixels per fingertip and 0.02N pressure sensitivity, supporting high success rates on long-horizon, complex manipulation tasks, for instance involving in-hand rotation and bimanual operation.
- Unitree H2 humanoid chassis, with 31 degrees of freedom across the body.
- NVIDIA Jetson AGX Thor T5000 onboard compute, featuring NVIDIA Blackwell architecture, an Arm CPU, 128 GB of unified memory, and a configurable power range.
Sharpa Wave: A Human-Like End Effector for Humanoid Development
With the Sharpa Wave mounted on the H2 Plus humanoid robot and simulated within NVIDIA’s platform, developers can:
- Capture high-quality upper-body demonstration data for dexterous manipulation training and policy development, in conjunction with NVIDIA Isaac Teleop.
- Simulate, train and evaluate dexterous and tactile manipulation policies using Sharpa Wave’s tactile parameters within Isaac Sim and Isaac Lab, NVIDIA’s open simulation frameworks, before committing to real-world deployment.
- Deploy dexterous and tactile manipulation policies in real time with minimal sim-to-real gap, leveraging NVIDIA Jetson for on-robot inference and control.
About Sharpa
Founded in 2024, Sharpa is a unicorn AI robotics company dedicated to developing high-performance dexterous robots, systems, and components for general-purpose applications. Sharpa’s mission is to build robots that free people from repetitive or strenuous work to focus on more meaningful pursuits. Sharpa’s global headquarters is in Singapore, with manufacturing R&D in Shanghai and business operations in Mountain View, USA.
