Industry
Zhejiang Humanoid Lands 2,000-Robot Apparel Order With Jack Technology
Zhejiang Humanoid Robot Innovation Center has signed a strategic partnership and 2,000-unit robot order with Jack Technology for garment manufacturing. The deal targets one of robotics' hardest industrial problems, handling soft fabric with millimeter-level accuracy at production scale.
Zhejiang Humanoid Robot Innovation Center and Jack Technology have signed an order for 2,000 customized NAVIAI robots for garment manufacturing, according to Gasgoo and Chinese technology press coverage published around the May 8 to May 11 window.
The order matters because apparel is not an easy factory sandbox. Fabric shifts, wrinkles, folds, and stretches. Zhejiang Humanoid says its system can keep multi-layer stacking alignment below 2 mm, while core sewing-related motions in the sector often require 0.3 to 0.5 mm precision.
AI-generated image
Flexible fabric handling is the technical core of the Zhejiang Humanoid and Jack Technology order. Source: AI-generated editorial illustration.
Key Stats
2,000
Robots Ordered
<2 mm
Stacking Accuracy
<10 sec
Single-Step Operation
2025
POC Validated
The Apparel Robot Order Story
The deal pairs Zhejiang Humanoid, a Ningbo-based humanoid robotics center, with Jack Technology, one of China's best-known industrial sewing equipment companies. Gasgoo described the agreement as a strategic partnership plus an order for 2,000 humanoid robots customized for garment manufacturing.
That framing is important. The most credible humanoid deployments in 2026 are usually in automotive plants, warehouses, or electronics factories. Those sites offer structured parts, repeatable bins, and clear workcells. Apparel is different. A sleeve, collar, denim panel, or polo shirt does not behave like a machined bracket. The material can sag under its own weight, catch on tooling, or change shape as soon as it is picked up.
Zhejiang Humanoid began specialized development for NAVIAI garment production tasks in September 2024. By July 2025, the company said the robot's template-machine workflow had passed proof-of-concept validation, covering opening and closing templates, fabric manipulation, stacking, and unloading.
Why template machines matter
Template machines hold fabric pieces in fixed patterns so sewing can be repeatable. Automating the feeding and unloading around those machines can remove a high-friction manual step without redesigning the whole factory.
Why soft goods are hard
Rigid parts can be localized with cameras and grasped against known surfaces. Fabric requires perception, force control, and motion planning that adapts as the object deforms.
Key Insight
The news is less about one robot model and more about the buying signal. A 2,000-unit apparel order suggests customers are starting to test humanoids where conventional automation has struggled for decades.
Under the Hood: Flexible Manipulation at Millimeter Scale
The technical bar in apparel work is unusually high. Gasgoo reported that fabric alignment deviations for pieces such as collars and pockets must be kept within plus or minus 2 mm, while cutting and sewing tasks can demand motion precision of 0.3 to 0.5 mm. Those numbers are much tighter than the tolerance needed to carry a tote across a warehouse aisle.
Zhejiang Humanoid says NAVIAI uses proprietary full-stack hardware and software, including an embodied intelligence sensing and control system for fabric perception and fine manipulation. The system is reported to handle fabrics from polo shirts to denim, complete single-step operations in under 10 seconds, and maintain multi-layer stacking alignment below 2 mm.
AI-generated image
The deployment target is not a general home environment. It is a constrained apparel production line with repeatable tools and measurable outputs. Source: AI-generated editorial illustration.
| Spec | NAVIAI Apparel Robot | Agility Digit | Figure 03 |
|---|---|---|---|
| Primary Target | Garment manufacturing | Logistics and material handling | Factory and home tasks |
| Reported Scale | 2,000-unit order | Commercial RaaS deployments | Production ramp at BotQ |
| Manipulation Focus | Soft fabric separation and stacking | Tote handling and transport | General object handling |
| Known Accuracy Metric | <2 mm stacking alignment | Task throughput metrics by site | Autonomy duration and task success |
| Commercial Status | Purchase order announced | Commercial deployments | Internal and partner deployment |
Technical Context
• Soft-object perception: Fabric changes shape during handling, so the robot needs feedback during the motion, not just a pre-grasp image.
• Force control: Too little grip loses the garment. Too much grip distorts the fabric or jams the template.
• Teleoperation data: Zhejiang Humanoid says remote teleoperation supports long-term data collection for garment workflows.
Who's Building It
Zhejiang Humanoid Robot Innovation Center is positioned as the robot developer behind NAVIAI. The center has been working on full-stack hardware and software for embodied AI, with this apparel project focused on repeatable production work rather than general consumer chores.
Jack Technology brings the factory channel. The company is tied to industrial sewing equipment and apparel automation, which gives the robot program access to real process knowledge. That matters because a robot that looks competent in a lab can still fail when it meets fabric dust, operator changeovers, mixed garment batches, and production takt time.
• Zhejiang Humanoid: Developed NAVIAI garment workflow R&D starting in September 2024 and reported POC validation in July 2025.
• Jack Technology: Sewing equipment company using the partnership to push deeper into intelligent apparel manufacturing.
• Aitu Intelligent Sewing: Jack-linked subsidiary reported in Chinese coverage as part of the company's AI sewing and robot strategy.
• Apparel factories: The end market is labor-intensive, high-mix, and difficult to automate with fixed-purpose machines alone.
Sep 2024
R&D Start
Jul 2025
POC Passed
May 2026
Order Reported
What This Means for Apparel Manufacturing
The order points to a practical route for humanoid robots: start in narrow but valuable workflows where factories already understand the process economics. Garment production has a long history of partial automation. Cutting, sewing, pressing, and material handling all have specialized equipment. The gap is the connective tissue between steps, especially when humans still align, separate, and load fabric by hand.
If NAVIAI can perform reliably around template machines, the economic case is clearer than a vague general-purpose robot pitch. The buyer can measure units per hour, error rate, rework rate, machine utilization, and human supervision load. Those numbers determine whether the robot becomes a production tool or an expensive pilot.
The labor context also matters. Apparel factories still rely on people for repetitive handoffs because the objects are cheap, variable, and hard to fixture. A robot that can load, align, and unload fabric around existing machines could raise equipment utilization without asking the factory to replace every sewing station at once. That makes the path less risky than a full factory redesign.
There is also a data advantage. Every corrected pick, failed separation, and successful stack becomes training material for the next software revision. In a high-volume garment line, that feedback loop can generate more useful manipulation data than a staged lab demo, because the robot sees real variation in fabric, lighting, batch size, and operator workflow.
The risk is that purchase orders do not automatically equal full deployment. A 2,000-unit order can be staged, conditional, or tied to performance gates. Readers should treat this as a serious commercial signal, not proof that thousands of robots are already sewing garments at full utilization.
Business Signal
Humanoid robotics is moving from broad claims toward vertical workflows. Apparel is a strong test case because success requires dexterity, perception, repeatability, and production economics in the same cell.
What's Coming Next
The next milestones are delivery schedule, acceptance testing, and published factory metrics. The most useful numbers would be uptime, garments processed per hour, intervention frequency, fabric damage rate, and the number of workers required to supervise each cell.
Watch whether the program stays confined to template-machine handling or expands into adjacent steps such as sorting, folding, inspection, and machine feeding. A narrow win is still valuable, but the broader apparel automation story depends on how many manual transitions the robots can remove.
Frequently Asked Questions
What did Zhejiang Humanoid and Jack Technology announce?
They announced a strategic partnership and an order for 2,000 customized NAVIAI robots for garment manufacturing. The robots are aimed at fabric handling around apparel production workflows, including template-machine processes.
Why is garment manufacturing difficult for robots?
Fabric is deformable, so it does not keep a stable shape like a metal or plastic part. Garment work often needs millimeter-level alignment, gentle force control, and the ability to handle different fabrics without constant reprogramming.
What performance numbers have been reported?
Coverage reported multi-layer stacking alignment below 2 mm and single-step operations under 10 seconds. The sector's core cutting and sewing tasks can require motion precision of 0.3 to 0.5 mm, which explains why this use case is technically demanding.
Are these robots already deployed in factories?
The public news describes an order and partnership, not a completed 2,000-unit rollout. The important follow-up will be delivery timing, factory acceptance tests, and measured production performance.
The 12-Month Outlook
The NAVIAI order gives China's humanoid sector a concrete apparel manufacturing story at a time when many robot announcements still lean on demos. If Zhejiang Humanoid and Jack Technology can publish hard production data, this could become one of the cleaner benchmarks for whether humanoids can solve soft-goods automation at scale.
The best outcome is not a robot that does everything. It is a robot that does a specific, painful factory task well enough to be ordered again. Apparel gives the industry a demanding test with real commercial stakes.
The Bottom Line: A 2,000-unit apparel order is a sharper signal than another viral walking demo because it ties humanoid robotics to measurable factory work.
For the next year, the question is simple: can NAVIAI move from purchase order to repeatable throughput on real garment lines?