On June 24, 2026, Xiaopai Technology held the “Brand Strategy Upgrade Conference and New Species Launch Conference” in Shenzhen, and officially released two core intelligent terminal product lines: the first home intelligent terminal ClawStation series, and the dedicated home care terminal “KinSight”, announcing that the smart home industry has officially entered a new stage of visual AI Agent development.
The ClawStation series released this time covers three types of use scenarios: the full-featured flagship home AI terminal ClawStation Pro, the ClawStation standard version that relies on cloud computing power to lower the hardware threshold, and the ClawStation exploration version that is installed on TV boxes in the form of an Android App, which can adapt to the intelligent transformation needs of different families.
Among them, ClawStation Pro, which is positioned as the home AI brain, has attracted the most attention. This product is equipped with a 7B parameter local large model and comes with independent local computing power. It can provide full-link “environmental perception” and “intelligent decision-making” support for the whole-house smart home system. It can also control different brands of smart home products through standardized interaction protocols to achieve cross-brand whole-house intelligent linkage.
In terms of core perception capabilities, ClawStation
Pro jumps out of the traditional smart home technology path: Most smart home systems in the past relied on infrared sensors and millimeter-wave radars to detect human presence. Such solutions do not require cameras and have low privacy risks, but they have obvious shortcomings – they cannot accurately identify users and can only track larger human targets. They have no ability to sense small items such as scattered garbage and water cups forgotten on the table, making it difficult to achieve truly active smart services.
ClawStation Pro uses a visual sensor solution as the core of perception, and can be connected to up to four home cameras. The AI analysis process of all video images is completed locally, and it can “understand” various events occurring in the home scene without uploading to the cloud.
Relying on this capability, the product unlocks many new functions that traditional smart homes cannot achieve: For example, it can accurately identify whether the person in the current space is an elderly person, a child or a visitor, and actively adjust the air conditioning temperature and light brightness according to the usage habits of different groups of people; when it recognizes scattered garbage on the ground, it can automatically issue cleaning instructions to the sweeping robot; when it detects dangerous scenes such as the elderly falling or children climbing the windowsill, it will also push warning information to the user as soon as possible.
In response to the visual perception privacy issue that users are most concerned about, ClawStation
At the press conference, Pro emphasized the core features of local analysis and edge computing: all video data is only transferred and analyzed within the local device of the home and will not be uploaded to any external server, thus avoiding the risk of leakage of family privacy data from the technical roots.
In addition, Xiaopai Technology also announced the opening of the camera device interface, providing standardized capabilities such as device search, security authentication, audio and video acquisition, and device settings. Users can build personalized smart scenes according to their own needs and further expand the boundaries of product use.
Currently, the ClawStation series of products are not yet open for purchase by ordinary consumers, and the actual recognition accuracy and intelligent linkage capabilities have yet to be verified by the market. However, from the perspective of industry development, this technical path of “local visual AI + whole-house perception center” provides a new direction worth exploring to solve the current industry pain points of “passive response and scene fragmentation” in smart homes. It is expected to promote whole-house intelligence from the stage of “people command equipment” to officially enter a new stage of “devices actively serve people”.
