GreenTrans (綠捷), a subsidiary of China Motor (中華車), unveiled its new-generation quadruped robots GT5X and GT3X at SEMICON Taiwan 2026CITE:E1. The company said the launch is meant to build Taiwan's own core robotics technology, with a target of reaching 100% local-content manufacturing by 2027CITE:E1.
GreenTrans's quadruped robot R&D team was transferred over from China Motor's original automotive technical teamCITE:E3. The team brings more than 15 years of experience in automotive embedded systems, whole-vehicle control, automotive-grade CAN BUS communication architecture, and advanced driver-assistance technologyCITE:E3. Two core components — the Autonomous-Mobile-Robot Control Unit (ACU) and the battery management system (BMS) — were both designed in-house by the teamCITE:E3.
For core computing and AI motion control, GreenTrans partnered with Enjoy Tech (恩嘉科技) to adopt NVIDIA's Jetson Orin platformCITE:E4. The two companies also use NVIDIA's Isaac Lab reinforcement-learning training environment to optimize the robots' movement strategies across complex terrainCITE:E4.
On the energy side, GreenTrans is working with Formosa Smart Energy (台塑新智能) to jointly develop a lithium iron phosphate (LFP) battery system for the quadruped robotsCITE:E5. By contrast, GreenTrans's existing inspection robots already carry a roughly 100% localization rate and run on a lithium ternary battery supplied by Formosa Plastics (台塑)CITE:E7. GreenTrans associate vice president Jiang Meng-chang (江孟璋) said the quadruped robots will switch to the LFP system developed with Formosa Smart Energy once they reach the 2027 localization targetCITE:E7.
GreenTrans is targeting 100% local-content manufacturing for its quadruped robots by 2027CITE:E2. Jiang Meng-chang said the quadruped line is expected to hit that 100% localization rate in 2027, paired with the Formosa Smart Energy LFP battery systemCITE:E7.
GreenTrans's inspection robots have already been deployed in semiconductor fab settings since the second half of 2026CITE:E6. The company said the quadruped robots are instead expected to reach commercial deployment only after the second half of 2027CITE:E9.
| Robot line | Localization target | Battery system | Deployment status |
|---|---|---|---|
| Inspection robot | ~100% (already reached) | Lithium ternary battery, Formosa PlasticsCITE:E7 | Deployed in semiconductor fabs since H2 2026CITE:E6 |
| Quadruped (GT5X / GT3X) | 100% by 2027CITE:E2 | LFP system, Formosa Smart EnergyCITE:E5 | Commercial deployment expected after H2 2027CITE:E9 |
GreenTrans said its main strategic focus is a smart-factory software-and-hardware integration platformCITE:E8. That platform covers autonomous mobile robot (AMR) integration systems and smart management systemsCITE:E8. GreenTrans also partners with semiconductor equipment makers to develop inspection robots for fab facility applicationsCITE:E8.
GreenTrans is running two localization tracks on different clocks and different chemistries: inspection robots are already near-fully localized with Formosa Plastics' lithium ternary batteries and are already working inside semiconductor fabsCITE:E7, while the quadruped GT5X/GT3X line trails by roughly a year, aiming for the same 100% localization figure only after adopting a separately developed LFP battery system with Formosa Smart EnergyCITE:E5. The gap between the inspection robot's current field deployment and the quadruped robot's post-H2-2027 commercial timelineCITE:E9 lines up with GreenTrans's own framing of its business as an AMR-and-management-platform operation first, with legged robotics as the newer, still-localizing product lineCITE:E8.
GreenTrans (綠捷), a subsidiary of China Motor (中華車), unveiled its new-generation quadruped robots GT5X and GT3X at SEMICON Taiwan 2026CITE:E1.
GreenTrans's quadruped robot R&D team was transferred over from China Motor's original automotive technical teamCITE:E3.
For core computing and AI motion control, GreenTrans partnered with Enjoy Tech (恩嘉科技) to adopt NVIDIA's Jetson Orin platformCITE:E4.
On the energy side, GreenTrans is working with Formosa Smart Energy (台塑新智能) to jointly develop a lithium iron phosphate (LFP) battery system for the quadruped r…
The most telling detail here isn't the quadruped form factor itself — it's that GreenTrans built the Autonomous-Mobile-Robot Control Unit and battery management system in-house, leaning on a team with over 15 years of automotive-grade control experience rather than assembling a robot from off-the-shelf modules. That engineering base explains why the two product lines are staged so differently: inspection robots already carry a lithium ternary battery from Formosa Plastics and are already deployed in semiconductor fabs at roughly 100% localization, while the quadruped GT5X/GT3X line still needs to pair NVIDIA Isaac Lab-trained locomotion with a newly developed LFP battery system before it can claim the same localization rate, with commercial deployment not expected until after the second half of 2027. The metric worth watching next is whether that 2027 target holds, since it depends on two separate supplier integrations — Enjoy Tech's Jetson Orin compute stack and Formosa Smart Energy's LFP system — landing on the same schedule.
Nvidia confirmed on September 3, 2026 that it agreed to buy Hugging Face for $12.93 billion, exactly $12,930,300,000, gaining the open-source AI hosting platform used by over 18 million developers. CEO Jensen Huang pledged the platform will stay open, with no Nvidia compute required to build on or deploy through it.
NVIDIA will subscribe US$3.5 billion of MediaTek's US$3.9 billion offshore convertible bond offering, the largest such issuance in Taiwan's capital market history, deepening cooperation in AI infrastructure, edge AI computing, and automotive platforms while marking NVIDIA's first major investment in a Taiwanese company.
Utility-scale solar has become one of the cheapest new power sources, with costs down roughly 90% since 2010, while 2024 deployment volume outpaced every other generation technology. Paired with record-low battery prices, solar-plus-storage is already powering an AI data center in Arizona — though it still falls short of full 24-hour dispatchable baseload.