Techman Robot has offered its newest developments within the space of humanoid robotics at Nvidia GTC 2026, together with a movement capture-based coaching method for its TM Xplore I platform.
The corporate additionally introduced a collaboration with j-mex, integrating wearable movement seize know-how into its robotics improvement workflow.
Movement seize used to coach humanoid robots
The system combines Techman’s TM Xplore I humanoid robotic with j-mex’s Moxi movement seize swimsuit and VR interface, permitting operators to regulate the robotic by means of real-time human motion.
In response to the businesses, customers sporting a VR headset and movement seize swimsuit can replicate detailed actions – akin to dual-arm greedy and object sorting – with the robotic in actual time.
The method is designed to generate high-quality coaching knowledge by capturing human movement straight, which may then be used to enhance robotic efficiency in real-world duties.
Techman says this technique helps quicker improvement of “sim-to-real” capabilities, the place robots skilled in simulated or managed environments can function extra successfully in dynamic bodily settings.
Shift towards data-driven robotic coaching
The usage of movement seize displays a broader shift in robotics towards data-driven coaching strategies, significantly within the improvement of so-called “bodily AI” programs.
By amassing giant volumes of real-world movement knowledge, builders can prepare robots to carry out extra complicated and variable duties with out relying solely on conventional programming methods.
“Future automation will now not be restricted to monotonous, repetitive duties, however will totally advance towards deep, clever collaboration outfitted with notion and considering capabilities,” stated Scott Huang, chief working officer at Techman Robotic.
Collaboration with movement seize specialist j-mex
j-mex says its inertial measurement unit (IMU)-based movement seize know-how permits exact monitoring of human motion, which may be translated into robotic actions.
“Via high-precision IMU movement seize know-how, we are able to precisely convert delicate human actions into high-quality coaching knowledge. We’re delighted to work with Techman Robotic to collectively speed up the sensible software of Bodily AI,” stated Tsang-Der Ni, CEO of j-mex.
Positioning inside bodily AI improvement
Techman Robotic says it’s specializing in integrating notion, decision-making, and movement management capabilities throughout each collaborative and humanoid robotics platforms.
The corporate describes its method as combining imaginative and prescient, reasoning, and execution inside a unified system, reflecting wider trade efforts to develop robots able to working in much less structured environments.
The TM Xplore I demonstration at GTC highlights ongoing work to bridge the hole between human habits and robotic execution, as corporations discover new methods to coach machines for real-world purposes.
