As a member of the NVIDIA Connect program, j-mex continues to advance Physical AI, Humanoid Robotics, and Motion Intelligence through the NVIDIA ecosystem.
During the NVIDIA GTC Taipei 2026 keynote, NVIDIA Founder and CEO Jensen Huang showcased the company’s rapidly expanding AI ecosystem, featuring organizations driving innovation across AI, robotics, cloud computing, and research.
Among the organizations highlighted was j-mex Inc., whose logo appeared in the NVIDIA GTC Taipei 2026 Ecosystem Showcase, demonstrating the company’s growing presence within the global AI and robotics ecosystem.
The recognition reflects j-mex’s continued commitment to advancing Motion Intelligence, Motion Capture, Teleoperation, and Physical AI, while contributing technologies that enable the next generation of intelligent robotic systems.
Advancing Innovation Through the NVIDIA Ecosystem
As a member of the NVIDIA Connect program, j-mex leverages NVIDIA’s AI and robotics technologies to accelerate innovation in Motion Intelligence and Human-to-Robot applications.
By integrating technologies including NVIDIA Isaac Sim, NVIDIA Isaac GR00T, and ROS 2, j-mex has established an end-to-end workflow spanning human motion capture, teleoperation, motion retargeting, simulation, and robot demonstration data generation.
This workflow enables developers and research organizations to transform high-quality human motion into robot-ready demonstration data, significantly reducing the time and complexity required to develop and train Physical AI systems.
AgileMaster: Accelerating Human-to-Robot Learning
As Physical AI and Humanoid Robotics continue to evolve, collecting high-quality human demonstration data has become one of the most critical challenges for Vision-Language-Action (VLA) models and imitation learning.
To address this need, j-mex developed AgileMaster – Humanoid Teleoperation Solution, an integrated platform designed to bridge human motion and robot intelligence.
Built upon nearly two decades of expertise in IMU-based motion capture technology, AgileMaster enables developers to capture, retarget, and transfer human motion into robot-executable actions with high fidelity.
Key capabilities include:
- Human-to-Robot Motion Retargeting
- Humanoid Robot Teleoperation
- High-Fidelity Motion Data Collection
- Robot Demonstration Data Generation
- ROS 2 Integration
- NVIDIA Isaac Sim Workflow
- Digital Twin Applications
AgileMaster enables robotics developers to efficiently collect high-quality demonstration data, accelerate AI model training, and shorten the development cycle for next-generation intelligent robots.
From Motion Intelligence to Physical AI
For nearly two decades, j-mex has specialized in MEMS inertial sensing, motion capture, and human motion analysis technologies.
Today, those core technologies have evolved into comprehensive solutions supporting Teleoperation, Motion Retargeting, Digital Twin, Robot Learning, and Physical AI.
Earlier this year, AgileMaster received the COMPUTEX 2026 Best Choice Award in the Robotics & Drones category, recognizing its innovation in Human-to-Robot technology and its contribution to the future of intelligent robotics.
Moving forward, j-mex will continue expanding its Motion Intelligence platform while leveraging the NVIDIA ecosystem to help researchers, developers, and enterprises accelerate the adoption of Physical AI and Humanoid Robotics worldwide.