YUNJI Technology signs a thousand-unit order for parent-child scenarios, and the scheduling system gains further market validation

Recently, "the first stock of robot service intelligent agents" YUNJI Technology (2670.HK) officially signed a strategic cooperation agreement with Shenzhen Zhongyi Group, reaching a procurement order of one thousand units for the UP composite polymorphic robot · Little Pony children's riding cabin. This is the world's first large-scale intelligent agent order implemented in parent-child consumption scenarios, marking another market validation of the company's robot service intelligent agent's mobility capabilities and scheduling system. Valuation Logic Restructuring in Parent-Child Scenarios The Little Pony robot is an AI innovative product launched by YUNJI Technology for public parent-child scenarios, integrating complete capabilities such as AI intelligent interaction, QR code payment, and safe riding, mainly targeting scenarios like airports, shopping malls, and high-end hotels. Public data shows that China's children's consumption market has exceeded 4 trillion yuan, while AI interactive children's devices have a significantly higher per-unit revenue and payment conversion rate in commercial scenarios compared to traditional entertainment devices. YUNJI Technology CEO Li Quanyin stated at the signing event: "From hotels to airports, shopping malls, and other public parent-child scenarios, we are not just creating a product, but validating a model: a system that can be reused across multiple scenarios. With standardized chassis, we can switch scenes at will, and every time we enter a new scene, our business boundaries expand." The core of this "model validation" lies in the qualitative change in the revenue structure. Traditional robot sales are a "one-time transaction," where revenue ends once the equipment is sold—YUNJI is rewriting this logic through the HDOS subscription model. According to the 2025 annual report, HDOS subscription customers surged by 925% year-on-year; AI digital system revenue reached 87.93 million yuan, accounting for 29.2% of total revenue, a year-on-year increase of 57.3%; intelligent agent application revenue reached 33.87 million yuan, a year-on-year surge of 194.1%. More notably, the Little Pony robot is equipped with QR code payment capabilities, which allows for continuous revenue expansion through new retail profit sharing in addition to service fees. For B-end customers, this means improved per-unit revenue and operational optimization; for C-end users, it signifies a brand new interactive consumption experience. As the business model shifts from "one-time hardware sales" to "sustainable cash flow engines," the valuation logic given to YUNJI Technology by the capital market will fundamentally reshape accordingly. Moat Validation—Dual Barriers of Chassis Reuse × Intelligent Scheduling The essence of the Little Pony robot is another "cabin swap" of the UP composite polymorphic robot platform—while the underlying standardized mobile chassis remains unchanged, the upper layer is replaced with a children's riding module, achieving free switching between scenarios such as hotel delivery, factory inspections, and parent-child riding through "one machine for multiple uses, time-sharing reuse." This "standardized chassis, freely switchable scenes" platform model constitutes YUNJI's first barrier: chassis reuse significantly reduces marginal costs and deployment cycles. On top of the chassis, the HDOS intelligent scheduling system builds a "software moat." Its architecture, driven by both embodied intelligence and disembodied intelligence, covers front-end AI voice perception, mid-end intelligent decision distribution, and back-end data accumulation and visualization. At the level of collective intelligence, HDOS uniformly schedules robot clusters across brands and scenarios, with official data showing an average of 32,000 units online daily in 2025, a year-on-year increase of 18.5%—massive operational data continuously feeds back into algorithm iteration, forming an "increasingly intelligent" evolutionary closed loop Chassis reuse reduces marginal costs, collective intelligence amplifies network effects, and builds a dual technological barrier. Behind this platform capability is high-intensity R&D investment and a full-stack patent layout. During the reporting period, YUNJI invested CNY 62.15 million in R&D, accounting for approximately 20.6% of its revenue. At the same time, YUNJI has completed a comprehensive iteration of the AI Agent Intelligent Diagnosis System (HQAS), constructing a voice intelligence system that covers the entire process of check-in, guest control, and service requests, supporting 23 languages and mixed recognition of Chinese and English, which has been scaled in thousands of hotels. It is worth mentioning that by the end of 2025, YUNJI has accumulated 997 authorized patents, participated in and completed 11 national and industry standards, covering the entire chain of perception, decision-making, and execution, continuously delivering underlying technological energy for customized services in various scenarios, and laying a solid foundation for expanding its share in the intelligent market. Continuous Growth Under Multi-Scenario Product Coverage In the hotel service intelligent agent track, YUNJI still maintains a leading position. By the end of 2025, YUNJI has deployed over 40,000 hotels globally, a 33% increase from 2024, maintaining the top market share in domestic hotel intelligent agents. Meanwhile, non-hotel scenarios are rapidly expanding. The latest annual report shows that YUNJI's factory scenario revenue soared by 393% year-on-year, while technology scenario revenue increased by 160.4%, with hospital deployments rising from 130 to about 200. According to MarketsandMarkets, "the global AI Agent market size is expected to grow from USD 5.26 billion in 2024 to USD 52.62 billion by 2030, with a CAGR of 46.3%." Domestically, in August 2025, the State Council officially issued the "Opinions on Deepening the Implementation of the 'Artificial Intelligence +' Action," clearly proposing a target of over 70% penetration rate for new generation intelligent terminals and agents by 2027. The dual resonance of policy and market opens up new imaginations for YUNJI's long-term growth space. The signing of a thousand XPeng vehicles verifies YUNJI's commercialization capability. From a single product to platform capabilities, from hardware sales to "subscription + profit sharing," from hotel scenarios to diverse spaces such as factories, hospitals, shopping malls, and airports. This is not only an iteration of YUNJI's own business model but also a microcosm of the Chinese service robot industry transitioning from "single-point tools" to "intelligent agent infrastructure." In this process, every step of YUNJI's exploration provides a verifiable sample for the commercialization of intelligent agents