Core Insights - Chery's Starway ET5 officially launched on November 28, featuring Horizon's HSD urban assisted driving system and the Journey 6P computing solution, marking the start of large-scale production for Horizon's HSD [1][2] Group 1: Product Launch and Technology - The Starway ET5 is the first vehicle globally to incorporate Horizon's HSD urban assisted driving system, which utilizes the Journey 6P chip and offers advantages such as low system latency, strong defensive driving capabilities, smooth control in both lateral and longitudinal directions, and continuous self-evolution [2] - Horizon has secured partnerships with over 10 automotive brands and more than 20 vehicle models for the HSD system, with notable models like Chery's Fengyun T9L and iCAR V27 showcased at the Guangzhou Auto Show [2] Group 2: Market Outlook and Financial Projections - The company anticipates significant growth in revenue, projecting operating income of 3.543 billion yuan, 6.010 billion yuan, and 8.007 billion yuan for 2025, 2026, and 2027 respectively, with corresponding net profits of -686.5 million yuan, -135.8 million yuan, and -82.3 million yuan [3] - The current price-to-sales (PS) ratios are estimated at 29.65 times, 17.48 times, and 13.12 times for the years 2025, 2026, and 2027 [3] Group 3: Strategic Partnerships and Global Expansion - The company is deepening ties with leading Tier 1 suppliers, aiming to gradually penetrate global markets, with Bosch's mid-level intelligent driving solution based on the Journey 6M already delivered at scale, and ZF planning to launch a coPILOT system based on the Journey 6P in China by 2026 [2] - The company’s subsidiary, Diguo Robotics, has launched the first integrated SoC computing kit RDK S100 in June and plans to release a high-performance platform RDK S600 (560 TOPS) in Q1 2026 [2]

Investment Rating - The report maintains a "Recommended" investment rating for the humanoid robotics industry [7]. Core Insights - The humanoid robotics concept index experienced a decline of 5.68% from November 16 to November 21, 2025, underperforming the CSI 300 index by 1.91 percentage points. However, year-to-date, the humanoid robotics index has increased by 57.83%, outperforming the CSI 300 index by 41.26 percentage points [2][12]. - Notable stock performances include Weichuang Electric with a weekly increase of 18.56%, while Fangyuan Co. saw the largest decline at 21.32% [17]. Summary by Sections Weekly Market Review - The humanoid robotics index fell by 5.68% during the specified week, while year-to-date performance shows a significant increase of 57.83% [12][2]. - Weichuang Electric and Fangyuan Co. had the highest weekly gains and losses, respectively [17]. Weekly Highlights Policy Developments - Hubei's provincial leadership emphasized the need to enhance scene applications, particularly in integrating "Beidou + humanoid robots" and "Beidou + artificial intelligence" [20]. - Chongqing's Economic and Information Technology Commission released a plan to accelerate the application of embodied intelligent robots in enterprise digital transformation projects [21]. Product and Technology Iteration - Figure AI announced that its second-generation humanoid robot, Figure 02, has participated in the production of 30,000 vehicles at BMW's Spartanburg plant, loading over 90,000 components [22]. - UBTECH's annual production capacity for industrial humanoid robots is projected to reach 5,000 units by 2026 [22]. - Diguo Robotics partnered with Fourier to enhance humanoid robots' capabilities in emotional perception and semantic understanding [23]. Investment and Financing - The embodied intelligence company Star Epoch completed nearly 1 billion yuan in Series A+ financing, led by Geely Capital [25]. Key Company Announcements - GAC Group plans to initiate large-scale production of humanoid robots by 2027, aiming for a production value exceeding 1 billion yuan by 2030 [26]. - Gree Electric has developed core components for humanoid robots, including joint modules [27]. - Shengshi Technology is advancing the localization of the NAO robot supply chain and is set to launch the NAO7 model soon [27].

北京商报讯(记者陶凤王天逸)11月19日，地瓜机器人宣布，与北京人形机器人创新中心有限公司达成基 于全新国产大算力平台RDK S600的深度合作。作为地瓜机器人S600全球首批合作伙伴，北京人形将在 其新一代具身天工机器人平台上，全面搭载地瓜机器人国产大算力平台RDK S600，共同推动具身智能 从追求极限运动能力的"最能跑"，向自主规划决策的"最聪明"，以及支撑复杂场景应用的"最好用"跃 迁。 ...

Core Viewpoint - The article discusses the transition from autonomous driving technology to robotics, highlighting the challenges and opportunities in the robotics industry, particularly in the context of embodied intelligence and the potential impact of new models like VGGT on 3D perception and robotics applications [5][7][60]. Group 1: Industry Trends - The robotics industry is at a pivotal moment, with significant technological advancements and a shift towards embodied intelligence, which is seen as the next frontier for AI [5][7]. - The article emphasizes the differences between the autonomous driving and robotics sectors, noting that while autonomous driving has reached a level of standardization, robotics is still exploring diverse hardware forms and algorithms [10][14]. - The VGGT model is introduced as a potential game-changer for 3D geometry, akin to how Transformers revolutionized natural language processing, indicating a shift towards unified solutions for 3D perception [6][67]. Group 2: Technological Migration - The migration of technology from autonomous driving to robotics is highlighted, with companies like DiGua Robotics leveraging experiences from the autonomous driving sector to enhance their robotics platforms [14][18]. - The challenges of hardware diversity in robotics are discussed, as the lack of standardization complicates data accumulation and algorithm development [10][14]. - The article outlines the evolution of autonomous driving algorithms from modular approaches to end-to-end systems, which are now being adapted for robotics applications [25][27]. Group 3: VGGT and Its Implications - VGGT is presented as a foundational model that could redefine 3D visual technology, offering a new paradigm for solving traditional geometric problems through large-scale data and models [55][67]. - The potential for VGGT to replace expensive depth cameras with cheaper RGB cameras is discussed, which could significantly reduce the cost of robotics systems [64][66]. - The article concludes that VGGT represents a significant advancement in the field of 3D vision, marking the entry of large models into the realm of geometric processing [67][68].