星动纪元亮相WRC：向全球展现具身智能「六边形战士」产品家族的实力

Core Viewpoint - The article highlights the impressive capabilities and commercial achievements of ROBOTERA's products showcased at the 2025 World Robot Conference, emphasizing their "soft and hard integrated full-stack self-research" in the field of embodied intelligence [3][19]. Group 1: Product Overview - ROBOTERA's product family, referred to as "Hexagonal Warriors," includes the Star L7, Star XHAND1, and Star Q5, which demonstrate comprehensive capabilities in both physical and cognitive dimensions [4][19]. - The Star L7 is the first full-size bipedal humanoid robot in China, showcasing its ability to perform complex movements like a 360-degree jump, highlighting its strength and stability [6][9]. - The Star XHAND1 features a fully direct-drive design with 12 degrees of freedom, capable of performing over 1,000 complex operations, making it suitable for various tasks [13][14]. - The Star Q5, designed for remote full-body operation, exhibits high flexibility and stability, capable of mimicking human movements seamlessly [16][18]. Group 2: Technical Capabilities - The Star L7's performance is comparable to Tesla's Optimus, with a peak torque of 400 N·m and a peak speed of 25 rad/s, allowing it to perform high-intensity tasks with human-like agility [9][10]. - The Star XHAND1 can exert a maximum grip force of 80 N, easily lifting 25 kg, and demonstrates rapid response times, akin to an esports player's speed [12][14]. - The Star Q5 supports real-time language switching across 37 languages, showcasing its adaptability in various service scenarios [18]. Group 3: Commercialization and Market Position - ROBOTERA has delivered over 200 units of its products and has numerous orders pending, with nine of the top ten global tech companies as clients, indicating its leading position in the embodied intelligence sector [19]. - The integration of the self-developed end-to-end VLA embodied model ERA-42 with hardware products creates a positive feedback loop for data collection, model training, and scene implementation, enhancing product performance [19].