Hangjiashuo Robot
On October 9, the China National Intellectual Property Administration announced the design patent for BYD's humanoid robot (authorization announcement number: CN310247768S). This marks the first official announcement of the design patent for "Xiao Di".
However, this is not the first time the product's appearance has been exposed to the public—the robot made its global debut with a real machine at "Di Space" in Zhengzhou in August 2026 and has since entered regular service.
Release Schedule: From the "Early August" Teaser to the Debut in August
On July 25, 2026, the Zhengzhou Di Space official account released a suspense poster stating, "In early August, a new friend wants to meet you," and BYD also confirmed the plan to release it at Di Space in August. However, on July 29, staff members stated that the event was still in a confidential stage, and the relevant promotional posts were subsequently deleted, leading the public to believe the release had "failed to materialize." Ultimately, however, Xiao Di still made its debut at Di Space within August.
Product Appearance: Patent Views Disclosed, But Not the First Exposure
The patent drawings show that Xiao Di adopts a gray and white color scheme, with a rounded curved surface design for the head and no prominent display screen; the torso, arms, and legs are equipped with a large number of joint structures, giving the limbs a high degree of freedom of movement.
The brief description of the patent states that the product is used for intelligent robots with walking, balancing, and interactive functions, and can be applied in service, education, or industrial scenarios.
Hardware and Perception: Core Components Highly Self-Developed
According to official and public information, Xiao Di stands 1.61 meters tall, weighs 58.5 kg, and features 31 degrees of freedom across its entire body. Its dexterous hands achieve a repeat positioning accuracy of ±1 mm and can grasp heavy objects weighing 1 kg. It integrates facial recognition, lip reading, and gesture recognition, supporting real-time mutual translation of 6 dialects and 6 foreign languages. Relying on 360-degree surround perception and dynamic modeling, it can autonomously avoid obstacles in complex exhibition hall environments and identify living beings and transparent objects such as glass. Over 80% of the core components are self-developed, mainly reusing Fudi Powertrain servo motors, Blade Batteries, and BMS, along with the 360° surround vision algorithms from the intelligent driving team.
Software and Self-Development Path: RobotOS and Vertical Large Models
At the software level, Xiao Di is equipped with BYD RobotOS 1.0, transplanting intelligent driving domain control algorithms with an end-to-end control latency of less than 50 milliseconds, and has built its own vertical AI large model for industrial scenarios. However, it still relies on external cooperation for motion control algorithms and perception solutions, and has not yet formed an end-to-end native self-developed embodied large model foundation.
Mass Production Plan: Factories First, Then Stores, and Finally Homes
Regarding the mass production plan, BYD plans to deploy the robots in its own factories within 2026, prioritizing the replacement of high-intensity repetitive positions on assembly lines. Subsequently, it will enter commercial use in stores, with a planned deployment of 2 to 3 units per dealer store to handle guest reception, vehicle model explanations, and in-car system demonstrations, expecting commercial batch implementation within one to two years. The long-term goal is to enter households, with the cost per unit controlled within CNY 200,000.
Industrial Synergy and Challenges: Ecosystem Emerging, Difficulties Remain
In terms of industrial synergy, BYD has signed a strategic cooperation agreement with Pacini Sensing Technology and invested in embodied intelligence enterprises such as AgiBot. From its August debut to the October patent announcement, Xiao Di's implementation path is gradually becoming clear. However, the depth of self-development in motion control and perception solutions, as well as its adaptability in complex scenarios, remain core challenges it must face.
This article is compiled based on publicly available network information and is for reference only, not constituting investment advice.