On the evening of March 29, Shenzhen IoT unicorn Lumi Lianchuang (Aqara) officially submitted its application to the Hong Kong Stock Exchange.
As a veteran player in Xiaomi’s ecological chain, this national-level specialized and new giant has been deeply involved in the smart home track for 17 years.
Unlike most smart home manufacturers, which bet on the success of a single product, Lumi has early anchored the “spatial intelligence” track in advance. Now, overseas revenue has surpassed domestic revenue, and gross profit margin has increased by nearly 9 percentage points in three years.
So, can Greenmi, which has acquired space intelligence in advance, really reap the first wave of dividends in the wave of smart home upgrades?
The predecessor of Lumi is Lutuo Technology, which was founded in 2009 by You Yanjun, who majored in automation. It initially engaged in the smart building business, which actually laid the foundation for the subsequent intelligent layout of space.
In 2014, Lumi changed its name to Lumi after joining Xiaomi’s ecological chain, and officially entered the C-end smart home track. While doing ODM manufacturing to accumulate production capacity experience, it also polished its own Aqara brand.
While the industry is still debating between single-product intelligence and interconnection, Lumi has already set its sights on spatial intelligence.
The spatial intelligence defined by Lumi is not as simple as connecting a bunch of devices, but based on sensors, hubs, and controllers, and through a unified semantic system, upgrading single-point device control to overall perception, decision-making, and automatic execution of spatial status and scenes.
This positioning coincides with the evolution of the smart home industry: from single-product intelligence to device interconnection, and now to active space intelligence, Lumi’s early layout has become the core technical barrier.
According to Frost & Sullivan, Lumi is the world’s largest Chinese space intelligent infrastructure provider in terms of revenue in 2024, with a global market share of 1.5%, ranking seventh.
Many people’s impression of Lumi is still that of “Xiaomi foundry”, but judging from the data disclosed in the prospectus in the past three years, Lumi’s proactive structural adjustments have begun to bear fruit.
From 2023 to 2025, Lumi’s overall revenue will be stable in the range of 1.4-1.5 billion yuan, the proportion of its own brand Aqara spatial intelligence business will increase from 62.6% to 64.3%, and the proportion of ODM foundry business will decrease from 36% to 33.8%.
As the proportion of self-owned brands increases, the overall gross profit margin rises from 29.7% to 38.4%, a cumulative increase of nearly 9 percentage points in three years, and profitability continues to be optimized.
What deserves more attention is the change in the regional income structure: calculated from its own brand business, Lumi’s overseas revenue in 2023 will be 380 million yuan, accounting for 42.2%; by 2025, overseas revenue will rise to 629 million yuan, accounting for 66.5%, and it will overtake domestic revenue in just two years.
The core focus of Lumi’s overseas growth is its differentiated advantages that are deeply tied to Apple’s ecosystem.
By the end of 2025, Lumi has more than 340 products connected to the Apple Home ecosystem, natively supporting all exclusive functions of Apple HomeKit, and its products have entered Apple retail stores in 18 countries and regions around the world.
In 2022, Lumi launched the world’s first smart door lock that supports Apple Home Key. The U400 smart door lock released at CES in 2026 is the first product on the market to support UWB technology for hands-free unlocking of Apple devices.
This strategy of deeply integrating top overseas ecology has helped Lumi quickly open up a gap in the mainstream European and American markets, and also allowed it to gain a differentiated ticket in the overseas competition of domestic smart home manufacturers.
The core difficulty of spatial intelligence has never been to stack hardware, but to solve the unified scheduling problem of different protocols and different ecological equipment.
Lumi’s technical idea is to build a unified software semantic layer on top of the three layers of hardware to solve compatibility pain points from the bottom layer.
Lumi has built an ontology framework called ORAP, which abstracts objects, relationships, actions and strategies in space into semantic models that can be understood by machines, and then synchronously maintains real-time data of spatial status through spatial maps.
Simply put, this framework allows the system to understand not “a certain light is on” but “someone enters the living room and the reading scene needs to be turned on”.
This architecture brings two very key advantages: First, cross-protocol compatibility. Lumi integrates more than ten protocols such as Zigbee, Wi-Fi, Matter, and KNX into a semantic model through capability mapping. The upper-layer logic does not need to perceive the differences in the underlying protocols to complete cross-device and cross-ecological scheduling.
The second is privacy and reliability. Human behavior recognition and vital signs monitoring at the perception layer are all processed locally, and the original data does not need to be uploaded to the cloud. The hub serves as a local computing node and can maintain automated operation even when the network is disconnected, solving the pain point of many smart home systems being paralyzed when the network is disconnected.
In terms of R&D investment, Lumi’s investment has increased from 173 million yuan to 234 million yuan in the past three years, and the R&D proportion has increased from 12% to 15.9%. The current R&D team has 460 people, accounting for nearly half of the total number of employees. It has accumulated more than 1,200 authorized patents, and its technical barriers continue to be consolidated.
The most discussed issue in the market is Lumi’s dependence on Xiaomi customers: Xiaomi has been Lumi’s largest customer in the past three years, and its revenue share has stabilized at more than 32%. Is this ratio considered high? Will it bring risks?
Judging from the data, this dependence is actually a two-way binding, and Lumi’s customer decentralization is already advancing: apart from Xiaomi, the second largest customer’s revenue accounted for less than 4%, and the top five customers’ combined share dropped from 42.7% to 38.9%, showing a continued trend of dispersion.
On the other hand, Xiaomi is not only a customer, but also a shareholder. It has invested in Lumi since the angel round. This ecological cooperation played a key supporting role in the early growth of Lumi.
From the supply side, Green Rice’s supplier dispersion is also increasing. The proportion of procurement from the top five suppliers has dropped from 21.7% to 19.3%, and the proportion of the largest single supplier has dropped from 7.3% to 4.9%. There is no obvious supply-side concentration risk.
In terms of ownership structure, the founding family holds approximately 39.27% of the voting rights through direct shareholding and voting rights agreements, maintaining control of the company. Xiaomi holds 7.92% of the shares, Shunwei Capital holds 8.37% of the shares, and the remaining shareholders are financial investors such as Joy Capital, Far East Horizon, and Cathay Innovation. The ownership structure is relatively stable.
As for the continuous losses shown in the prospectus, they are actually mainly accounting losses caused by changes in the fair value of convertible and redeemable preferred shares, and do not represent actual operating losses. This is a common financial phenomenon before the IPO of many technology companies.
Today, the smart home industry has gone from competing for subsidies and single product scale in the early days to a new stage of competing for system capabilities and scenario experience. Spatial intelligence is the core battlefield in the next stage.
When the Matter protocol gradually becomes popular, the basic conditions for the interconnection of multi-ecological devices are already in place, and users’ demand for “active intelligence” is getting stronger and stronger. Manufacturers that can provide complete spatial intelligence infrastructure will become the most valuable link in the industry chain.
Spatial intelligence is not a competition for a certain hit product, but a competition for the underlying architecture and ecological compatibility. The winner in this field must be the player who has built the framework, accumulated computing power, and laid out channels in advance.
The biggest direction of Lumi’s IPO investment is to continue to invest in R&D and upgrade its spatial intelligence technical capabilities. The second is to expand sales channels and brand building to further amplify the first-mover advantage it has gained.
The penetration rate of spatial intelligence in the domestic smart home market is still very low, and overseas markets are increasingly accepting Chinese smart hardware. Can Greenmi, which has been stuck in advance, take advantage of its IPO and turn its technological advantages into market success?
The answer will take time, but the explosion of this track may come sooner than most people expect.
