Core Viewpoint - The focus of competition in China's new energy vehicle sector is shifting from electrification to intelligence, particularly with the recent approval of L3-level conditional autonomous driving models, paving the way for future L4-level Robotaxi operations [1] Market Potential - By 2030, the market size for Robotaxi in China is expected to reach 158.3 billion yuan, with an estimated fleet size of 400,000 vehicles [1] - Currently, the Robotaxi market has not yet experienced a large-scale explosion, and commercial monetization is still in the exploratory phase [3] Company Overview - Xiaoma Zhixing (Pony.ai) was founded in 2016 and focuses on providing autonomous driving technology and solutions for transportation and logistics [5] - The company has established R&D centers in multiple locations, including Silicon Valley, Beijing, and Guangzhou, and aims to expand its product and business layout in Europe, the Middle East, and Asia [5] Financial Performance - In the first three quarters of 2025, Xiaoma Zhixing reported revenues of $60.88 million, a year-on-year increase of 54.1%, but a net loss of $157 million, widening from a loss of $93 million in the same period of 2024 [10] - The company's R&D expenses reached $157 million in the first three quarters of 2025, representing 257.74% of its revenue [12][13] Business Segments - The main sources of revenue for Xiaoma Zhixing in the first three quarters of 2025 were from autonomous driving trucks (43%) and technology licensing (39%), while Robotaxi services accounted for only 10% of total revenue [13] Competitive Landscape - The global Robotaxi industry has evolved into a multi-dimensional competitive landscape, with players like Tesla and Xiaoma Zhixing focusing on integrated models, while traditional ride-hailing platforms leverage user data [5] - Xiaoma Zhixing's unique technology moat is its "world model" (PonyWorld), which can simulate 10 billion kilometers of testing scenarios weekly, significantly enhancing AI training efficiency [6] IPO and Market Reaction - Xiaoma Zhixing's recent dual listing on NASDAQ and the Hong Kong Stock Exchange is seen as a necessary step to secure funding amid ongoing operational losses [7][8] - The stock price of Xiaoma Zhixing fell by 9.28% on its first day of trading, reflecting market concerns over the long-term profitability of L4 autonomous driving companies [10] Future Outlook - Xiaoma Zhixing plans to use approximately 50% of the net proceeds from its IPO to advance its market entry strategy and promote the large-scale commercialization of L4 autonomous driving technology over the next five years [15]

文｜王欣 编辑｜苏建勋 在2025年火热的具身智能创业潮中，"它石智航"有着绝对吸睛的实力。 这是一个由国内智驾黄埔军校核心高管组成的"梦之队"。它石智航首席执行官陈亦伦曾在华为车BU担 任自动驾驶系统CTO；首席科学家丁文超曾是华为"天才少年"。董事长李震宇则担任过百度智能驾驶事 业群原总裁，打造过全球最大的Robotaxi出行平台"萝卜快跑"。 在自动驾驶行业，陈亦伦、李震宇均是带过千人团队、打过胜仗的"名将"，两人的合作创业，也让它石 智航迅速成为资本的宠儿。在今年3月，它石智航以1.2 亿美元的融资额，创下中国具身智能行业天使轮 最大融资额纪录。 资本看重它石智航的技术积累和人才储备。线性资本创始人兼 CEO 王淮曾这样评价它石智航："他们能 将之前在华为做自动驾驶的很多软硬件打磨的经验，结合大模型的思考和推理能力，落实在具身机器人 身上。" 可在天使轮融资破纪录，创始团队如此豪华的状况下，不同于其他具身智能公司高频地披露出货量与技 术突破，2025年一年，它石智航鲜少公布进展。 12月19日，它石智航办了一场线上发布会，持续时间只有短短40分钟，展示的成果，是"全球首个完成 刺绣的机器人"。 为什么 ...

Core Insights - The article discusses the development of ColaVLA, a unified visual-language-action framework for autonomous driving that enhances trajectory planning by leveraging cognitive latent reasoning and hierarchical parallel planning [4][10][50]. Group 1: Background and Challenges - Traditional autonomous driving systems separate perception, prediction, and planning into distinct modules, while recent end-to-end (E2E) systems integrate these tasks into a unified learning pipeline [3][6]. - Visual-language models (VLMs) have been increasingly integrated into autonomous driving systems to inject cross-modal prior knowledge and world knowledge, but they face three core challenges: modal mismatch, high latency from autoregressive reasoning, and inefficiencies in planner design [7][9]. Group 2: ColaVLA Framework - ColaVLA proposes a unified framework that shifts the reasoning process from explicit text-based chains to a unified latent variable space, combined with a hierarchical parallel trajectory decoder [10][18]. - The cognitive latent reasoning component efficiently completes scene understanding and decision-making through two forward propagations, extracting decision-relevant information from multimodal inputs [11][21]. - The hierarchical parallel planner generates multi-scale trajectories in a single forward pass, maintaining causal structure and significantly reducing reasoning latency [12][28]. Group 3: Experimental Results - ColaVLA achieved state-of-the-art performance on the nuScenes benchmark, with the lowest average L2 error of 0.30 meters and a collision rate of 0.23%, outperforming existing action-based methods [37][38]. - In closed-loop evaluations, ColaVLA reached a NeuroNCAP score of 3.48, significantly improving safety metrics by reducing average collision rates from 65.1% to 36.8% [39][40]. - The framework demonstrated over five times the reasoning speed compared to traditional text-based autoregressive models, showcasing its efficiency and robustness [40][41].

Core Viewpoint - The article emphasizes the establishment of a comprehensive community for autonomous driving knowledge, aiming to facilitate learning, sharing, and collaboration among industry professionals and newcomers in the field [22][23]. Group 1: Community and Learning Resources - The "Autonomous Driving Heart Knowledge Planet" has been created to provide a platform for technical exchange, academic discussions, and engineering problem-solving, with members from renowned universities and leading companies in the autonomous driving sector [22][23]. - The community has over 4,000 members and aims to grow to nearly 10,000 in the next two years, offering a rich environment for both beginners and advanced learners [8][10]. - Various learning resources, including video tutorials, articles, and structured learning paths, are available to help members quickly access information and enhance their skills in autonomous driving [10][16]. Group 2: Technical Insights and Developments - Recent updates include insights from industry leaders on topics such as Waymo's latest base model, advancements in self-driving technology, and discussions on data loops and training cycles [7][10]. - The community has compiled over 40 technical routes covering various aspects of autonomous driving, including VLA benchmarks, multi-modal models, and data annotation practices [10][23]. - Members can engage with industry experts to discuss trends, technological advancements, and challenges in mass production of autonomous vehicles [11][26]. Group 3: Job Opportunities and Career Development - The community provides job recommendations and internal referrals to help members connect with potential employers in the autonomous driving industry [16][26]. - Regular discussions on career paths, research directions, and practical applications in the field are facilitated to support members in their professional growth [25][96]. - The platform encourages collaboration and networking among members, fostering a supportive environment for career advancement [20][26].

Domestic News - The State-owned Assets Supervision and Administration Commission (SASAC) emphasizes accelerating the transformation and upgrading of traditional industries, focusing on intelligent, green, and integrated development to enhance the competitiveness of traditional industries globally [6] - The National Energy Administration calls for the construction of a high-quality charging infrastructure system, aiming for a "three-year doubling" action plan to improve service capabilities and safety in electric vehicle charging [7] - In November, China's electric vehicle exports surged by 87% year-on-year, reaching 199,836 units, with significant growth in Asia (71%) and Latin America (283%) [8] - Changan Automobile plans to raise no more than 6 billion RMB through stock issuance to enhance its R&D and market competitiveness in the new energy vehicle sector [9] - NIO has announced the launch of a battery swap station in Deqin, completing a battery swap route over 2,700 kilometers with 19 stations [12] - CATL has established a new energy technology company in Xishuangbanna, focusing on new energy technology research and sales [13] International News - The market share of imported cars in South Korea is expected to exceed 20% in 2025, driven by the surge in Tesla sales and high demand for premium brands [15] - TIERIV has received support from the Japanese government to build a large-scale autonomous driving data set, enhancing industry competitiveness through AI data collection [16] - Waymo is conducting autonomous taxi tests in London, showcasing its electric vehicles on the streets [17] - General Motors plans to launch its sixth-generation V8 engine in 2027, offering two displacement options of 5.7L and 6.7L [18] Commercial Vehicles - The first fully vehicle-compliant unmanned logistics vehicle is set to launch in early 2026, with several domestic manufacturers focusing on unmanned logistics solutions [21] - Dongfeng Motor has completed dual national standard certification for vehicle information security and software upgrades, positioning itself at the forefront of compliance in the industry [23] - Iveco's electric vehicle has been awarded the title of "Green and Efficient City Distribution Recommended VAN" at a recent industry event [24] - The Fei Die Qitu MX (air brake version) light truck has been launched in Shaanxi, expanding the product lineup to meet diverse transportation needs [26]

转自：中国证券报·中证网 值得注意的是，易控智驾所布局的"不持车"轻资产模式，使得公司营收与活跃矿卡数量已全面反超传统 的"持车"模式，公司不仅获得了紫金矿业、宁德时代、兖矿资本、德赛西威等产业巨头的战略投资与深 度合作，更在商业化落地与技术创新层面持续突破，展现出强劲的发展潜力。 车队规模及营收双增 更新后的招股书显示，易控智驾在运行的无人矿卡车队规模从6月份的1400余台增长到2000台以上，公 司已占据中国L4级矿区无人驾驶解决方案市场的近半份额。截至2025年9月末，公司解决方案已部署在 中国按年核定产能计12个最大露天煤矿中的7个。截至最后可行日期，公司已于中国西北地区两个露天 矿山运输项目分别部署了509台和488台无人矿卡，刷新了行业最大规模无人矿卡车队纪录。 从收入规模看，进入2025年，易控智驾基本面持续优化。2025年前三季度，公司实现收入9.21亿元，同 比增长103.76%；毛利达到6532.8万元。 业绩攀升背后，是易控智驾商业模式的成功蜕变：公司从持车模式向不持车模式的商业化转型，正带领 矿区无人驾驶产业从"资产运营"到"技术赋能"深刻转向。 一般来说，目前矿区无人驾驶解决方案主要 ...

Core Insights - The global autonomous driving industry is entering a high-quality development phase by 2025, focusing on technological iteration, commercialization, and globalization [1] - The leading autonomous driving technology company, WeRide, is expanding its business coverage and steadily advancing the large-scale operation of core projects, achieving significant results in global projects [1] - WeRide has won 62 authoritative awards both domestically and internationally, gaining recognition from various media and government agencies [1] Global Expansion - The Middle East is a key area for WeRide's globalization efforts in 2025, transitioning from pilot projects to large-scale commercial operations [3] - In November, WeRide partnered with Uber to launch the first L4 level fully autonomous Robotaxi service in Abu Dhabi, marking the first city outside the U.S. to offer this service on the Uber platform [3] - The initial operation covers key areas like Yas Island, with plans to expand citywide, and a fleet of nearly 150 vehicles is expected by the end of 2025, with over 100 Robotaxis [3] Southeast Asia Development - In September 2025, WeRide's first batch of Robotaxi GXR models began testing in Singapore, completing multiple rounds of simulation road tests to lay the groundwork for M1 certification and public road deployment by year-end [5] - Previously, WeRide launched the first L4 level autonomous driving minibus route in Singapore and initiated the first commercial autonomous sanitation project, creating a multi-scenario landscape [5] Future Outlook - WeRide is outlining a clear path for the globalization of autonomous driving through benchmark projects, showcasing its robust capabilities and advancing the development of global smart transportation [7] - The company aims to continue its core logic of "technology-driven, commercial landing" with more mature solutions and broader coverage to benefit more regions and populations [7]

智通财经APP获悉，博雷顿(01333)尾盘涨超5%，截至发稿，涨2.81%，报25.6港元，成交额2368.9万港 元。 在近期举行的发布会上，博雷顿正式推出了面向无人驾驶场景正向开发的专用矿卡。从结构到线控系 统，从电驱平台到整车控制逻辑，完全以无人化作业需求为起点进行重构，标志着博雷顿将无人驾驶视 为未来矿山装备的底层能力，而非额外附加的功能。此前，博雷顿董事长陈方明在接受专访时表示，从 2026年起，博雷顿将大力推广无人驾驶、智慧矿山业务。 消息面上，近日，希迪智驾正式登陆港交所，引起业界对自动驾驶电动矿卡赛道的广泛关注。资料显 示，该公司专注于封闭环境自动驾驶电动卡车、V2X(车联网)技术及智能感知解决方案的研发等。行业 数据显示，2024年中国无人驾驶电动矿卡出货量约1400辆，预计到2026年将增至5500辆。业内普遍已将 2025年视作矿山无人驾驶电动卡车"规模化应用起量之年"。 ...

中国商报（记者 王彤旭）2025年，面对复杂多变的内外部发展环境，我国民营经济在政策赋能与自身突围中乘风破浪、韧性彰显。年终岁尾回望，从民营 经济促进法正式实施筑牢法治根基，到各地因地制宜精准施策；从创新赛道上的加速领跑，到外贸市场中的顶压前行，民营经济以全方位的突破，成为稳定 宏观经济大盘、推动高质量发展的重要支撑。 数据显示，截至2025年5月底，全国实有民营经济组织达1.85亿户，占经营主体总量的96.76%，其中民营企业数量突破5800万户，同比增长5.2%，民营经 济"主力军"作用愈发凸显。民营经济正以稳健的发展姿态，书写着新时代的成长答卷。 截至目前，浙江、山东、山西、海南等多地已出台促进民营经济发展的新政。如海南结合自贸港特色制定《海南自由贸易港促进民营经济发展若干规定》， 从保障公平竞争等方面出发，旨在通过一系列制度设计，打通当地民营经济发展的痛点难点。又如山东省财政厅印发《进一步支持民营经济高质量发展若干 法治筑基、政策协同：构建全链条支持体系 "传播民企法治理念，实现守法普法目的。"12月29日，东莞市观音山森林公园开发有限公司（以下简称观音山公司）董事长黄淦波在微信朋友圈分享了一篇 题为 ...

Core Insights - The article discusses two core papers from Horizon Robotics: DiffusionDrive and ResAD, focusing on their contributions to end-to-end autonomous driving solutions [2][3]. DiffusionDrive - The overall architecture of DiffusionDrive consists of three parts: perception information, navigation information, and trajectory generation [6]. - Perception information includes dynamic/static obstacles, traffic lights, map elements, and drivable areas, emphasizing the need to convey perception tasks to planning tasks in an end-to-end manner [6]. - Navigation information is crucial for avoiding incorrect routes, especially in complex urban environments like Shanghai, where navigation challenges are significant [7]. - The core concept of trajectory generation is "Truncated Diffusion," which leverages fixed patterns in human driving behavior to reduce training convergence difficulty and inference noise [8][10]. - The article outlines a method for trajectory generation using K-Means clustering to describe common human driving behaviors, which simplifies the training process [9]. ResAD - ResAD introduces a residual design that predicts the difference between future trajectories and inertial extrapolated trajectories, rather than generating future trajectories directly [12]. - The residual regularization helps manage the increasing residuals over time, ensuring that the model focuses on the true diversity of driving behaviors [13][14]. - The design allows for different noise perturbations in the trajectory generation process, adjusting learning difficulty based on the direction of motion [15]. - ResAD also features a trajectory ranker that utilizes a transformer model to predict metric scores based on top-k trajectory predictions and environmental information [16]. Conclusion - Both papers from Horizon Robotics provide valuable insights and methodologies for enhancing autonomous driving systems, encouraging further exploration and development in the field [18].