SHENZHEN, ChinaJune 11, 2026 /PRNewswire/ — When a resident of Chenzhou, Hunan Province walks into a China Mobile service hall, they may not just pick up a SIM card or a router. Instead, they could receive a set of robot data collection equipment. After a brief training session, they will return home as a “data collector,” completing collection tasks in their daily lives and providing the scarce real-world operational data that robots need.
World’s First ‘Embodied Data Collection 5S Store’ Launched: Daimon and China Mobile Build ‘Data Collection into Homes’ Network
Recently, Daimon Robotics and China Mobile officially partnered to launch the “Data Collection into Homes” network initiative. Leveraging China Mobile’s tens of thousands of offline retail outlets, the two parties will jointly build a nationwide outsourced data collection network. The first outsourced data collection joint innovation pilot base has been established in Chenzhou, Hunan, and is scheduled to begin regular operations on July 15.
Positioned as an “Embodied Data Collection 5S Store,” the base integrates five core modules: Show (display), School (data collection training), Supply (data collection equipment provision), Service (pre-sales and after-sales support), and Synergy (data-model-scenario collaboration). This initiative aims to transform data infrastructure from closed-door collection to large-scale, publicly accessible supply.
Embodied Data Collection 5S Store
Solving Embodied AI’s “Data Challenge” with the Real World
The generalization capability of embodied AI models is rooted in vast interactive experiences within real-world environments. However, the industry has long faced a common bottleneck: traditional data collection factories have high operational costs, relatively simple and homogeneous tasks, and often produce repetitive data, making it difficult to support the training needs of general-purpose models for “diversity, complexity, and long-tail scenarios.”
Operational data from the real world is precisely what best meets the training needs of embodied models. Take home environments, for example—they are among the most complex real-world settings, with countless variations in objects, random changes in lighting and space, frequent task interruptions, and significant individual differences in human operations. This is exactly the kind of real-world data that embodied models urgently require.
Daimon’s approach is to “outsource” data collection from fixed locations, extending it into real-world scenarios such as homes, manufacturing, services, and research labs, allowing models to observe a broader, more varied, and more realistic physical world.
Solving Embodied AI’s ‘Data Challenge’ with the Real World
In April of this year, Daimon released the world’s largest multimodal physical world dataset with tactile feedback, Daimon-Infinity, and pioneered an outsourced data collection system to fundamentally address the shortcomings of traditional data collection factories.
Now, Daimon’s collaboration with China Mobile further scales this approach: turning the “real world” into a sustainable data source and making “public participation” a replicable network capability.
Collect, Learn, and Use Instantly: Service Halls Become “Data Collection Gateways”
As the world’s largest telecom operator, China Mobile’s extensive network of offline outlets provides a broad and stable foundation for the data system.
These offline outlets, serving as “Embodied Data Collection 5S Stores,” offer the public one-stop training services for instant collection, learning, and use. They are equipped with various collection devices, including two-finger grippers, five-finger gloves, and head-mounted cameras, covering five core scenarios: home, logistics, manufacturing, exhibition halls, and retail. After short-term training, citizens can become data collectors, entering real-world scenarios to complete collection tasks, thereby maximizing the authenticity and diversity of human operations.
The “Data Collection into Homes” project starts in Hunan and will gradually expand to other regions across the country, ultimately forming an outsourced data collection network anchored by offline outlets, with surrounding residential households as the smallest units, covering all aspects of daily life. The initial phase of the project will deploy 1,000 sets of equipment, capable of generating 1 million hours of data annually at full capacity. Through standardized collection processes accessible to the public, it will rapidly accumulate highly scarce real-world operational data for the industry.
This partnership not only solidifies the foundation of a “national-level supply network” for Daimon’s physical world data ecosystem but also marks a critical step for China Mobile’s deep involvement in embodied AI infrastructure construction.
Upgrading China Mobile’s vast offline service hall network into a frontline for tactile data collection
From Investment to Co-Construction: Daimon × China Mobile Accelerates General-Purpose Embodied AI Deployment
Notably, this collaboration is the first tangible business outcome following China Mobile’s Chain Leader Fund investment in Daimon Robotics and the participation of China Mobile’s Embodied AI Industry Innovation Center in the Daimon-Infinity dataset. It signifies an extension of the partnership from capital and R&D to infrastructure co-construction.
Duan Jianghua, Founder and CEO of Daimon, stated: “The core competition in embodied AI has now shifted to the continuous acquisition of high-quality real-world data. Our collaboration with China Mobile allows us to push data collection from closed-door factories into real homes. This will be the first piece of infrastructure in the era of embodied AI and a step the entire industry has been waiting for.”
Looking ahead, Daimon and China Mobile will work together to build a large-scale data collection network covering hundreds of cities and connecting billions of people. Every real-world operation and tactile sensation will become a cornerstone for bringing general-purpose embodied intelligence to reality.
