Toborlife AI launches full-body teleoperation software for Unitree G1

Toborlife AI introduced U.S.-built teleoperation and training software for the Unitree G1 Edu humanoid on Sept. 18, giving labs and developers a domestic stack for data capture, support and onboarding. The system is designed to let teams start recording training-ready demonstrations from the first session while keeping customer data on U.S.-controlled hardware.

Why it matters: - Toborlife AI’s software gives U.S. research labs, universities and developers a domestic teleoperation and training layer for the Unitree G1 Edu humanoid. - The system is built to capture training-ready data from the first session, which can shorten setup time and speed experimentation. - Customer data stays on the customer’s hardware in the United States and is never transmitted to Unitree or any server outside the customer’s control.

What happened: - Toborlife AI announced full-body teleoperation software for the Unitree G1 Edu humanoid on Sept. 18, 2026. - The company described the release as the first data collection, training and support stack for the G1 built and supported domestically in the U.S. - The software is available for purchase now with U.S.-based onboarding and support included. - Customers can schedule a consultation or request a live demonstration by contacting sales@toborlife.ai or (408) 409-4061.

The details: - The system lets an operator wearing a PICO VR headset move the G1 Edu in real time, with no scripting in between. - Full-body teleoperation covers the robot’s arms, legs and torso. - Optional dexterous robotic hands on the G1 let operators pick up, hold and manipulate objects with finger-level control. - The software records full-body motion at high fidelity, including finger-level hand movement. - Toborlife AI delivers the captured demonstrations in a training-ready format rather than raw logs that need cleaning. - The company backs deployments with a U.S.-based team of robotics engineers who provide direct technical support, configuration guidance and hands-on onboarding. - George Ekas, co-founder and CTO, said the end-to-end data collection and training system is designed to support domain-specific applications and improve as customers use it.

Between the lines: - Toborlife AI is positioning teleoperation as infrastructure, not the final product. - The pitch is aimed at teams that want to start collecting usable robot data immediately instead of spending months integrating open-source tools. - Data ownership is a central selling point. The company says customer data stays with the customer, is not monitored, and is only accessed for troubleshooting or support when requested. - That model could appeal to labs and companies that want to build proprietary skill libraries and training sets without giving up control of their demonstrations. - David Schulhof, CEO and founder, said the product is a milestone in the company’s effort to democratize emerging robotics and make robots easier to use for people without deep technical backgrounds.

What's next: - Toborlife AI is taking orders now and offering live demos for research groups and companies that want to evaluate the system. - The company’s support model suggests it will continue pairing software updates with onboarding and optimization for deployed customers. - For labs building humanoid workflows, the next step is likely moving from single demonstrations to repeatable training pipelines that can be validated on the G1 platform.

The bottom line: - Toborlife AI is betting that the next robotics race is not just about hardware, but about who owns the software, data and support layer around humanoid robots.