Lingbao CASBOT Releases its First Full-Sized Bipedal Humanoid Robot "CASBOT01", Ushering in the Commercialization Year of Humanoid Robots
Lingbao CASBOT Releases its First Full-Sized Bipedal Humanoid Robot "CASBOT01", Ushering in the Commercialization Year of Humanoid RobotsOn November 13th, Lingbao CASBOT, a humanoid robot brand, officially unveiled its first full-sized bipedal humanoid robot, "CASBOT01." This multi-scenario application-oriented general-purpose brain-like intelligent robot marks a significant step forward for Lingbao CASBOT in the field of humanoid robots and signals the official arrival of the commercialization era of humanoid robots
Lingbao CASBOT Releases its First Full-Sized Bipedal Humanoid Robot "CASBOT01", Ushering in the Commercialization Year of Humanoid Robots
On November 13th, Lingbao CASBOT, a humanoid robot brand, officially unveiled its first full-sized bipedal humanoid robot, "CASBOT01." This multi-scenario application-oriented general-purpose brain-like intelligent robot marks a significant step forward for Lingbao CASBOT in the field of humanoid robots and signals the official arrival of the commercialization era of humanoid robots.
"CASBOT01" stands tall at 179cm, weighs 60kg, boasts 52 degrees of freedom, possesses a computing power of 550T, and has a battery life exceeding 4 hours. Through numerous optimizations, Lingbao CASBOT has achieved a balance between range of motion and aesthetic appeal in the mechanical structure design of "CASBOT01." With extreme weight reduction throughout the machine, high-strength metallic materials are used to process complex curved surfaces, resulting in a super-thin hollow cavity with reinforcement in the middle, ensuring structural stability and impact resistance.
To achieve highly efficient and stable motion control, "CASBOT01" has independently developed three series of integrated joints: planetary, harmonic, and linear. These joints have a peak torque density of up to 207Nm/kg and a joint efficiency exceeding 80%. Lingbao CASBOT also utilizes a force control optimization strategy based on current or sensors, supplemented by model tracking control. Furthermore, technologies like vibration suppression, inertia identification, friction compensation, and "gain self-tuning" are employed to achieve "high-precision torque loop" control effects. Through inertia matching, the dynamic response frequency of individual joints is enhanced. Efficient and stable communication meets the high bandwidth requirements for dynamic force control. Impact scenario optimized energy management strategy, combined with servo, single-leg, and full-machine overcurrent protection measures, has completed nearly 1,000 hours of life testing, ensuring the long-term stable operation of the joints.
The overall motion control framework of "CASBOT01" combines antagonistic motion priors and whole-body control (WBC), enabling integrated control of movement. In complex dynamic environments, the robot can execute whole-body operational tasks with stability. Building upon the traditional reinforcement learning training framework, Lingbao CASBOT concurrently trains a generative adversarial imitation learning network. The generator and discriminator engage in continuous game-playing, allowing the model to generate highly human-like motion control strategies. In terms of model training, Lingbao CASBOT adopts an innovative technical path that progresses from real to virtual, data synthesis, hybrid training, and back to virtual to real. Through hierarchical end-to-end models and hybrid training techniques, the robot becomes more intelligent, flexible, and autonomous. When the "physical body" and "brain" work closely together, the robot gains high-level cognitive, comprehension, and decision-making planning capabilities.
The head of "CASBOT01" has two degrees of freedom and is equipped with multiple RGBD cameras, lidar, and IMU sensors, enabling both visual and auditory interactions. Under the operational framework, through end-to-end multi-modal dexterous operation large models, distillation and quantization techniques, the robot's hand has successfully achieved generalization and adaptability to "environment, object, and task."
As the end effector responsible for executing fine-grained operations, "CASBOT01" has a biomimetic five-finger dexterous hand weighing 800g with a rated payload of 5kg. It incorporates a highly compact integrated structure and a biomimetic design integrating drive, transmission, sensing, and control systems. Additionally, it features a multi-source sensing system design incorporating tactile, force, and visual perception. By optimizing control algorithms and motion planning, imitating the flexibility of a human hand during operation, it is aimed at tackling the challenges of "dexterity" and "task execution" in fine-grained operation scenarios. With strong software and hardware support, rapid learning and generalizable long-time-series dexterous manipulation are realized.
2024 is widely regarded as the "Commercialization Year of Humanoid Robots." A recent report released by the China Humanoid Robot Industry Conference indicates that the Chinese humanoid robot market will reach 2.76 billion yuan in 2024, with the potential to grow into a market worth hundreds of billions of yuan by 2030. Behind these figures lies a dual-sided reality in the humanoid robot industry: on the one hand, its robust industrial penetration and integration capabilities make it a crucial force in "new-type productivity" and a new pillar driving China's high-quality economic development. On the other hand, amidst a market worth hundreds of billions, commercialization emerges as a vital "question" for the humanoid robot industry.
At present, Lingbao CASBOT has cultivated numerous partnerships in multiple fields, including aerospace, maritime, emergency rescue, underground work, industrial manufacturing, and commercial services, exploring new application scenarios for embodied intelligence. With the release of "CASBOT01," Lingbao CASBOT will further deepen the application of humanoid robots in various domains, contributing to the advancement of the humanoid robot industry.
Disclaimer: The content of this article is sourced from the internet. The copyright of the text, images, and other materials belongs to the original author. The platform reprints the materials for the purpose of conveying more information. The content of the article is for reference and learning only, and should not be used for commercial purposes. If it infringes on your legitimate rights and interests, please contact us promptly and we will handle it as soon as possible! We respect copyright and are committed to protecting it. Thank you for sharing.(Email:[email protected])