Beijing robotics company Galbot is launching its first outlets outside mainland China with three humanoid-operated convenience stores in Hong Kong. South China Morning Post reported on August 31 that the stores, following an opening ceremony, were scheduled to launch on September 1 in Hung Hom, Kai Tak, and Wan Chai.
The locations give Galbot a public test of humanoid retail in a market that is commercially distinct from the mainland. SCMP reports that Galbot began with its first mainland outlet in August 2025 and has since built a network of roughly 200 locations distributed among 50 Chinese cities. Hong Kong becomes a first overseas-style step for the Beijing-based embodied-intelligence company, which plans to add about 10 more stores in high-footfall districts in coming years.
Three Stores Put Humanoid Retail in Daily Use
The September 1 launch brings Galbot into retail environments where customers can encounter machines in ordinary shopping routines rather than at a technology exhibition. The three sites are in Hung Hom, Kai Tak, and Wan Chai, districts that give the company varied urban locations for its first Hong Kong presence. SCMP did not provide technical specifications for the humanoid systems, store staffing levels, sales volumes, or details of the product range.
That means the launch should not be confused with proof of a fully automated retail model. The report establishes the locations, the scheduled opening, and Galbot’s expansion plan. The commercial questions remain open: how reliably the robots perform, how customers respond, how much human support is required, and whether store economics work at scale.
Galbot Chief Strategy Officer Zhao Yuli told SCMP that the Hong Kong operation is intended as a launchpad for international-facing business. The company has had a Hong Kong office since 2024, she said. That gives the rollout a strategic role beyond the three stores. It lets Galbot test a visible commercial format in a city that connects mainland technology companies to international capital, visitors, and business partners.
The experiment also follows a broader interest in robots as part of consumer strategy. EastFrontier’s August 16 report on robot shopping in Beijing examined how a Chinese city is presenting retail interaction as an AI application. Galbot’s Hong Kong shops move that idea from a municipal promotional setting toward a branded company rollout.
Hong Kong Offers an Application, Research, and Capital Setting
Hong Kong Financial Secretary Paul Chan said at the opening ceremony that the city can make a unique contribution to robotics in three areas: it can provide viable application scenarios, a sustainable research-and-development ecosystem, and a robust supporting capital chain. SCMP reported those remarks alongside the store opening.
The formulation explains why the city is attractive to a company like Galbot. A robotics firm needs more than an engineering lab. It also needs places to deploy machines, feedback from real users, partners that can operate sites, and financing for a long commercialization cycle. Convenience stores offer a measurable setting because tasks are repetitive, customer interactions are frequent, and the service environment is public.
Hong Kong has recently expanded AI tools in its own public-service environment. EastFrontier’s August 29 coverage of the city’s 1823 platform explored a government use case for automated inquiry handling. Galbot’s stores present a different version of the same policy question: how AI systems can operate where people expect a dependable service, a clear route to help, and accountable management when something goes wrong.
SCMP describes Galbot as an embodied-intelligence robotics company. The phrase points to systems that interact with physical surroundings instead of operating only in software. In a retail store, that physical interaction involves navigating a public space and serving customers in conditions that are less controlled than a factory or laboratory. The success of the Hong Kong rollout will depend on whether its systems can sustain that activity day after day.
From Opening Ceremony to Commercial Evidence
Galbot’s mainland record provides a starting point. The company says it has expanded from its first store to about 200 outlets across 50 cities since August 2025. The Hong Kong plan adds three initial stores and a stated intention to open about 10 more. The scope is meaningful, but the reported numbers are company expansion claims rather than independently verified measures of store profitability or robot uptime.
The next evidence will come from operations. Do customers use the stores frequently? What tasks do the robots perform without intervention? How many employees support each location? How are safety, product availability, accessibility, and maintenance handled? SCMP’s report does not answer these questions, and neither should an early-stage deployment be judged by a launch event alone.
Still, the choice of Hong Kong makes the experiment worth watching. A successful rollout would show that a mainland Chinese robotics company can adapt a public-facing format outside its home market while using Hong Kong as a commercial bridge. A difficult rollout would be equally informative, revealing where humanoid retail remains too expensive, unreliable, or constrained by the realities of service work.
Galbot’s three stores are therefore not a verdict on humanoid commerce. They are an operating test. The company is creating a real-world setting in which a prominent physical-AI ambition can be evaluated by customers, landlords, investors, and regulators in real time.