How Zhejiang Humanoid Is Commercializing NAVIAI Robots

Zhejiang Humanoid Robot Innovation Center is building a commercialization case around specific industrial jobs rather than the promise of a robot that can do everything a person can do. Founded in Ningbo at the end of 2023 by Professor Xiong Rong and her Zhejiang University-linked team, the company is taking NAVIAI into manufacturing, logistics and commercial settings where a customer can measure throughput, precision and labor impact. 4348

The commercial model: sell a workflow, not a spectacle

The company’s clearest industrial targets are material handling and sorting, precision assembly, and garment manufacturing. Its product lineup reportedly includes fixed-base, wheeled-arm and bipedal variants—an important distinction because a work cell that does not require walking may not need the cost and complexity of a full biped. 2237

That approach is commercially pragmatic. A fixed system can be tailored to a repeatable station, while mobile or bipedal forms make more sense only where factory layout or the task genuinely requires mobility. The value proposition is therefore tied to a defined job: reliably manipulating objects, loading and unloading equipment, or handling materials that conventional automation struggles with.

Reports say NAVIAI systems have been introduced at industrial customers including SUPCON, GREE, JACK and FOTILE. Those references are meaningful as evidence of customer access, but public reporting does not provide enough independently verified detail to treat all of them as scaled fleet deployments. 42

Why garment handling is the central test

Garment production is Zhejiang Humanoid’s strongest near-term wedge. Flexible fabric wrinkles, collapses and changes shape during handling, making it far more difficult to automate than rigid parts. Xu Xuecheng, the company’s CTO and deputy general manager, has described recognition and manipulation of deformable cloth as a core challenge. 37

NAVIAI has been developed for template-machine operations such as opening and closing templates, manipulating fabric, stacking pieces and unloading material. The company said its template-machine solution passed proof-of-concept validation in July 2025. 1136

For garment workflows, published reports cite the following performance results:

• 98% fabric-piece separation success;
• about three seconds for a reported operation;
• ±2 mm positioning accuracy; and
• approximately 80% of human efficiency in the cited workflow. 13437

These metrics make the strategy more concrete than a generic humanoid demonstration. They should still be read carefully: the available reports do not establish independently audited performance across varied factories, shifts, fabric types or maintenance conditions.

The 2,000-unit Jack Technology program: major signal, incomplete proof

In April 2026, Zhejiang Humanoid and intelligent sewing-equipment company Jack Technology announced a strategic partnership and a signing for 2,000 customized garment humanoid robots. Multiple reports described it as the first large-scale humanoid application project in the apparel sector. 1540

It is a consequential announcement because it connects a robot developer with an established industry specialist and a narrowly defined production task. Yet the headline number should not be confused with 2,000 robots already operating in factories.

Public accounts conflict. Some reports characterize the arrangement as a signed 2,000-unit order and later describe it as entering batch delivery. 3536 Another report, citing Jack’s investor-platform explanation, says the 2,000 robots are a strategic-cooperation framework target still in technical development and scenario validation, rather than completed commercial delivery. 32

The most defensible reading is that the Jack arrangement is a large announced commercial commitment and a critical commercialization milestone, but not independently verified evidence of a 2,000-robot installed base. Whether it progresses into sustained deliveries will be the company’s most important proof point.

Building evidence at home, then expanding abroad

Zhejiang Humanoid’s international activity appears designed to establish references and gather operating feedback before broader overseas sales. Reporting indicates that one NAVIAI model has been deployed at a European home-appliance manufacturer’s plant, while the company has also presented NAVIAI logistics applications with KION at LogiMAT 2026 and appeared at AUTOMATE 2026 in North America. 2019

These steps matter, but they represent different levels of commercial maturity. A trade-show demonstration can validate interest; a pilot can test integration; a long-running production deployment is the stronger evidence of repeatable revenue. They should not be treated as interchangeable.

The company’s stated sequencing—refining products in Chinese facilities connected to multinational customers before expanding further into Europe and North America—fits that reality. The initial use cases are repetitive, labor-intensive and potentially hazardous jobs, where a buyer can assess performance using measurable operational outcomes rather than humanlike appearance. 37

What NAVIAI must prove next

Zhejiang Humanoid is also exploring commercial and retail applications. Reports describe NAVIAI in a “smart store manager” role capable of retrieving goods from shelves, alongside industrial use cases. 20 Those settings may broaden data collection and deployment experience, but factories remain the clearest route to near-term validation.

The company’s broader technology claims—such as a layered intelligence and control architecture—may help explain its product direction, but commercial success will be decided on the factory floor. The decisive questions are:

1. Reliability: Can robots sustain performance across long production runs rather than short demonstrations?
2. Integration: Can they work safely with existing machines, software and human operators?
3. Economics: Do deployment, maintenance and support costs justify the labor and quality gains?
4. Repeatability: Can a successful garment workstation be replicated across lines and customer sites?

Zhejiang Humanoid’s strategy captures a wider shift in humanoid robotics: from impressive hardware demonstrations toward constrained tasks, identified customers and measurable deployment milestones. The Jack program is the clearest test of that model. If it turns into a dependable production fleet, it would strengthen the case for task-specific humanoid automation; if it remains in validation, it will still show how far the sector must travel before announced volume becomes realized revenue.