TeleAvatar Teleoperation Body and TeleDroid System

Industry Classification
Complete machine development, design, system integration Humanoid robot
联系二维码

Scan the code to enquire about the product

Product Description

DexTeleop Intelligence's core product is the Embodied Intelligence Data Platform composed of the TeleAvatar Teleoperation Body and the TeleDroid System. Defined as the "most efficient data collection mother machine " this product aims to build the data cornerstone for embodied intelligence bridging the transmission pipeline from human skills to robotics. Defined as the "most efficient data collection mother machine " this product aims to build the data cornerstone for embodied intelligence bridging the transmission pipeline from human skills to robot skills. It addresses the critical industry pain point of the "data drought " specifically It addresses the critical industry pain point of the "data drought " specifically the lack of high-quality long-sequence complex operation data required to move embodied AI from L2 to L4 autonomy. The product consists of two main components: the TeleAvatar (the physical body) and TeleDroid (the hard-software-cloud integrated system). The TeleAvatar is designed specifically for teleoperation with extreme followability featuring a human-like configuration that allows for zero- and zero-threshold hands-on (the hard-software-cloud integrated system). The TeleAvatar is designed specifically for teleoperation with extreme followability featuring a human-like configuration that allows for zero-reshold hands-on (hands-on means getting started/operation). The TeleDroid system manages the entire data lifecycle including collection cleaning and automated annotation on the cloud side ensuring a closed-loop management of data assets. Key highlights of the product include "True Remote Teleoperation" versus the industry's standard "Near-field Visual Range Remote Control. Remote Control." While competitors require operators to stand within 1-2 metres of the robot DexTeleop enables operators to control robots from over 2 000 kilometres away in a comfortable and comfortable environment. While competitors require operators to stand within 1-2 meters of the robot DexTeleop enables operators to control robots from over 2,000 kilometers away in a comfortable central control room. This capability allows for "Space-Time Folding " enabling low-cost labour in regions like Guizhou to serve high-cost scenarios in Shanghai and provides safety isolation for hazardous environments like bomb disposal or nuclear maintenance. Technologically the product boasts extreme low latency. It achieves an end-to-end cross-city latency of under 100ms and a force/position hybrid control response of under 30ms. It achieves an end-to-end cross-city latency of under 100ms and a force/position hybrid control response of under 30ms. To ensure operator comfort and data quality it features binocular wide-angle hardware-level frame synchronisation which completely eliminates motion sickness. A major differentiator is the "Algorithm-Defined Hardware" strategy. By using mature mass-produced industrial components instead of expensive customised laboratory parts, the "Algorithm-City latency of under 100ms and a force/position hybrid control response of under 30ms. By using mature mass-produced industrial components instead of expensive custom laboratory parts the Bill of Materials (BOM) cost is only 1/3 of similar competitors. Despite the lower cost high precision is maintained through advanced algorithms such as high-precision calibration achieving 0.1mm repeat positioning accuracy. The product is designed to solve the "Sim-to-Real Gap." Unlike data from simulations or structured labs which produce low-entropy data DexTeleop collects high-fidelity data from the simulations or structured labs. collects high-fidelity data from unstructured real-world scenarios (complex lighting non-standard objects dynamic interference). This produces high-entropy data essential for training robust world models. Primary application industries include: Data Collection: Serving global This produces high-entropy data essential for training robust world models. Primary application industries include: Data Collection: Serving global universities AI research institutes and data collection centers needing high-quality embodied intelligence data. Automating loading and unloading tasks in complex environments. Unmanned Retail: Managing convenience stores and gas stations where one operator can Unmanned Retail: Managing convenience stores and gas stations where one operator can remotely manage 5-10 terminals (1:N multi-machine time-sharing). Unmanned Retail: Managing convenience stores and gas stations where one operator can remotely manage 5-10 terminals (1:N multi-machine time-sharing). MaaS (Manipulation as a Service): Evolving from L0 teleoperation to L4 general embodied intelligence where robots become infrastructure like electricity. The business model operates on a dual-engine drive: hardware sales (TeleAvatar Pro/Lite) as the cash cow and MaaS capability subscription as the growth pole. The roadmap progresses from L0 (Teleoperation 1:1 human-robot ratio) to L2 (Hybrid Augmented Intelligence 1:10 ratio) and finally to L4 (General Embodied Intelligence 1:10 ratio). L4 (General Embodied Intelligence <1:1000 ratio) creating a data flywheel that reinforces core data and intelligence barriers.

Product Specifications

The TeleAvatar and TeleDroid system delivers extreme performance metrics designed to bridge the gap between human dexterity and robotic execution. The performance specifications are categorized into latency precision force control kinematics hardware reliability and visual immersion. Responsiveness: The system achieves industry-leading low latency crucial for true remote teleoperation. The end-to-end cross-city network latency is maintained under 100ms. The end-to-end cross-city network latency is maintained under 100ms. Under ideal network conditions this can be further optimized to below 150ms for video return ensuring real-time visual feedback. Under ideal network conditions this can be further optimised to below 150ms for video return ensuring real-time visual feedback. More critically the force/position hybrid control response time is less than 30ms. This millisecond-level response is vital for maintaining the "feel" of the operation allowing the operator to react instantly to environmental changes without the lag that typically causes operational errors This millisecond-level response is vital for maintaining the "feel" of the operation allowing the operator to react instantly to environmental changes without the lag that typically causes operational errors or motion sickness in remote scenarios. The repeat positioning accuracy reaches 0.1mm ensuring that the robot can return to the exact same spot consistently for repetitive tasks. The repeat positioning accuracy reaches 0.1mm ensuring that the robot can return to the exact same spot consistently for repetitive tasks. The absolute positioning accuracy is maintained at 1mm. algorithms that correct for mechanical tolerances in standard hardware allowing the robot to perform delicate assembly or manipulation tasks that This level of precision is enabled by high-precision calibration algorithms that correct for mechanical tolerances in standard hardware allowing the robot to perform delicate assembly or manipulation tasks that usually require expensive high-end industrial arms. Utilising advanced current loop algorithms the system enables ordinary motors to exhibit " flexible touch." This allows the robot to be used for mechanical tolerances in standard hardware allowing the robot to perform delicate assembly or manipulation tasks that usually require expensive high-end industrial arms. Utilising advanced current loop algorithms the system enables ordinary motors to exhibit " flexible touch." This allows the robot to handle fragile objects such as picking up an egg without breaking it or manipulating soft materials where force feedback is critical. This allows the robot to handle fragile objects such as picking up an egg without breaking it or manipulating soft materials where force feedback is critical. This software-defined force control significantly reduces hardware costs while maintaining high sensitivity to weight texture and hardness properties of interaction targets. The robot employs a 7-axis mechanical arm configuration that mimics the human arm's degrees of freedom (20+ DOF overall for the embodied system). Through a global inverse kinematics algorithm the system eliminates singular points that typically restrict robot movement. This grants the TeleAvatar flexibility that surpasses that of a human arm. TeleAvatar flexibility that surpasses human hands in certain ranges allowing it to reach around obstacles and operate in complex spatial configurations that would be awkward or difficult to achieve. This grants the TeleAvatar flexibility that surpasses human hands in certain ranges allowing it to reach around obstacles and operate in complex spatial configurations that would be awkward or impossible for standard 6-axis arms or human operators in constrained spaces. Hardware Reliability and Cost Efficiency: The hardware strategy focuses on "Algorithm-Defined Hardware " selecting mature mass-produced components verified by the industrial sector rather than custom laboratory-grade parts. The hardware strategy focuses on "Algorithm-Defined Hardware" selecting mature mass-produced components verified by the industrial sector rather than custom laboratory-grade parts. This results in a Bill of Materials (BOM) cost that is only 1/3 of similar competitors. The modular design ensures maintenance is simple akin to changing a light bulb supporting 7x24-hour continuous operation. The components are proven through millions of industrial cycles ensuring high reliability and durability in harsh environments. Visual Immersion and Synchronization: To ensure the Visual Immersion and Synchronisation: To ensure the operator feels truly present ("TeleAvatar") the system features binocular wide-angle hardware-level frame synchronization. This technology completely eliminates motion sickness a common issue. completely eliminates motion sickness a common issue in VR and remote operation. The stereo vision provides depth perception essential for 3D spatial tasks. The immersive return allows the operator to rely on binocular stereo vision rather than naked eyes achieving a "being there" experience even when controlling a robot 2 000 kilometres away. The immersive return allows the operator to rely on binocular stereo vision rather than naked eyes achieving a "being there" experience even when controlling a robot 2 000 kilometres away. In terms of data performance the system is capable of collecting high-entropy data from unstructured real-world scenarios. Unlike simulation data which lacks complex contact friction and flexible deformation details TeleAvatar captures full-modal data including vision force touch and full-body posture. This data is isomorphic mapped meaning how the human moves is exactly how the robot moves making the data available to the user. This data is isomorphic mapped meaning how the human moves is exactly how the robot moves making the data the "gold standard" for training world models and simulators. Scalability: The system supports a scaling model from 1:1 (one operator one robot) to 1:10 (one operator managing 5-10 terminals via backend management in L2 hybrid mode). This scalability is supported by the low-latency architecture and the semi-autonomous fallback mechanism where the machine runs autonomously 95% This scalability is supported by the low-latency architecture and the semi-autonomous fallback mechanism where the machine runs autonomously 95% of the time and switches to teleoperation only for corner cases ensuring uninterrupted service guarantees.

Company Profile

Beijing Lingguo Intelligent Technology Co.

Spirit Royal Intelligence is a global leading teleoperated robot and data platform enterprise that defines the cornerstone of body intelligence data, founded by Tsinghua hardcore army. The company's main business is TeleAvatar teleoperated ontology and TeleDroid software-hardware-cloud integration system, which creates "the most efficient data collection mother machine", opens up the transmission pipeline from human skills to robot skills, and solves the pain point of "data shortage" of embodied intelligence. Adopting the strategy of "algorithm-defined hardware", the BOM cost is only 1/3 of similar competitors, with industrial-grade reliability and consumer-grade ease of use. The business model is double-engine driven: hardware sales (cash cow) + MaaS capability subscription (growth pole). In terms of honours and qualifications, the company has been featured in CCTV-2/4, 2026 CES has won the first overseas order, and scientific research and digital procurement has been certified by many customers. The company has won 7 championships in the remote operation track of the 2025 Zhongguancun Embodied Intelligence Application Competition, and the award of China Innovation and Entrepreneurship Competition (the first Embodied Intelligence Professional Competition), and has been invited to co-construct the national "Trustworthy AI - Embodied Intelligence Benchmarking Test Task Bank". The company's vision is to liberate human beings from dangerous, heavy and boring work, accumulate atomic action data through L0 teleoperation, and eventually evolve to L4 generalised embodied intelligence, so that robots can become an infrastructure like electricity.
Visitor Registration Call us WeChat

Leave Messages

Exhibitor Application
Visitor Registration Contacts Add WeChat

Reserve a Booth

Exhibitor Application

Medium/EN