Core Insights - The company has established a strong competitive moat in technology, licensing, mass production capabilities, and operational capacity, with significant advancements in autonomous driving technology and regulatory approvals [1][2] Group 1: Technology and Innovation - The company has developed a universal technology platform called WeRide One, which integrates AI models, software algorithms, modular hardware design, and cloud architecture, showcasing high versatility and scalability [1] - The recent launch of WeRide GENESIS has significantly enhanced the company's autonomous driving algorithms' ability to handle complex real-world scenarios, improving iteration efficiency and reducing the time and cost of traditional road testing [1] Group 2: Licensing and Regulatory Approvals - The company has achieved a first-mover advantage in licensing, having obtained autonomous driving licenses in eight countries, including China, UAE, Singapore, France, the USA, Saudi Arabia, Belgium, and Switzerland, by 2025 [1] Group 3: Fleet and Operational Capacity - As of January 12, 2026, the company's global Robotaxi fleet has reached 1,023 vehicles, with approximately 200 units deployed in the Middle East, aiming for tens of thousands of Robotaxis by 2030 [1] - The company is actively expanding its presence in the domestic Robotaxi market, with operations in major cities like Beijing, Guangzhou, and Shanghai, and plans to introduce autonomous taxi services in Shanghai by July 2025 [2] Group 4: Global Expansion Strategy - The company has a robust global expansion strategy, with plans to increase its Robotaxi fleet in the Middle East from 200 to 500-1,000 vehicles by 2026, and is set to launch a fully autonomous commercial operation in Abu Dhabi with Uber by November 26, 2025 [2] Group 5: Market Position and Future Prospects - The company employs a platform strategy in its L4 business, leveraging its success in autonomous taxi development to explore other autonomous vehicle solutions, including minibuses and freight vehicles [2] - By August 2025, the company is projected to hold a 36% market share in the global autonomous minibus market, ranking first based on revenue [2] Group 6: Financial Projections - The company forecasts net losses of 1.69 billion yuan, 1.37 billion yuan, and 1.10 billion yuan for the years 2025 to 2027, with corresponding EPS of -1.65 yuan, -1.33 yuan, and -1.07 yuan [3]

Investment Rating - The report initiates coverage with a "Buy" rating for the company [3][6]. Core Insights - The company has established a strong competitive moat in L4 autonomous driving through technology, licensing, mass production capabilities, and operational efficiency. It has developed the WeRide One platform, which integrates AI models, software algorithms, modular hardware design, and cloud architecture, enhancing scalability and versatility [1][28]. - The company has achieved significant milestones in licensing, having obtained autonomous driving licenses in eight countries, including China, UAE, Singapore, France, the USA, Saudi Arabia, Belgium, and Switzerland, positioning it as an industry leader [1][36]. - As of January 12, 2026, the company's global Robotaxi fleet reached 1,023 vehicles, with a target of deploying tens of thousands by 2030 [1][41]. Summary by Sections Company Overview - The company, WeRide, was founded in 2017 and is the only technology firm with autonomous driving licenses in eight countries. It focuses on developing safe and reliable autonomous driving technologies across various applications, including Robotaxi, Robobus, Robovan, and Robosweeper [16]. Technology and R&D - The company emphasizes R&D, with a high proportion of its workforce dedicated to it. As of June 30, 2025, 93.7% of its full-time employees are R&D personnel, including top AI scientists and engineers [28]. - The WeRide GENESIS simulation model significantly enhances the efficiency and capability of autonomous driving algorithms, reducing the time and cost of traditional road testing [31][32]. Licensing and Market Presence - The company has a clear advantage in licensing, having secured multiple licenses in various cities, including Shanghai and Abu Dhabi, where it has commenced fully autonomous operations [2][39]. - The company has successfully launched its Robotaxi services in major cities, including Beijing, Guangzhou, and Abu Dhabi, with plans to expand further [2][45]. Financial Projections - The company forecasts net losses of CNY 1.69 billion, CNY 1.37 billion, and CNY 1.10 billion for the years 2025, 2026, and 2027, respectively, with expected EPS of -CNY 1.65, -CNY 1.33, and -CNY 1.07 [3][5]. Market Strategy - The company is focused on both domestic and international markets, with a robust strategy for global expansion, particularly in the Middle East, where it has established a significant presence [2][82]. - The company aims to achieve single-vehicle breakeven in Abu Dhabi, indicating a viable business model for its Robotaxi services [54][86].

Core Insights - The development of intelligent driving is not a linear technological curve but a result of the interplay between various technical paradigms, engineering constraints, and real-world scenarios [1] - As the industry moves beyond the proof-of-concept stage, single technical terms can no longer explain the real differences in capabilities [2] - Factors such as computing power, data quality, system architecture, and engineering stability are determining the upper and lower limits of intelligent driving [3] Group 1: Evolution of Learning Techniques - Recent discussions in intelligent driving technology reveal a trend where various paths, such as end-to-end, VLA, and world models, converge on the concept of reinforcement learning [5] - Reinforcement learning is transitioning from a "technical option" to a "mandatory option" in the industry [7] - The emergence of products like AlphaGo and ChatGPT has highlighted the effectiveness of allowing AI to learn through trial and error as the fastest evolutionary method [8][9] Group 2: Learning Methodologies - Understanding reinforcement learning requires a grasp of imitation learning, which was previously favored in intelligent driving [11] - Imitation learning allows AI to learn from human driving data but has limitations, such as inheriting bad habits and struggling with unfamiliar situations [14][16] - Reinforcement learning, as demonstrated by AlphaGo, allows AI to explore new strategies through self-play, leading to superior performance beyond human intuition [17] Group 3: Reinforcement Learning Mechanisms - Reinforcement learning operates on a trial-and-error basis, where the model learns to drive well through a cycle of feedback [26] - The design of reward functions is crucial, as it translates driving performance into quantifiable scores [30] - Balancing conflicting objectives, such as safety versus efficiency, is essential in reward function design [32] Group 4: World Models and Advanced Learning - The integration of world models with reinforcement learning enhances the training environment, allowing AI to simulate real-world scenarios [42][49] - High-fidelity virtual environments enable AI to consider long-term consequences of actions, improving decision-making [50] - The coupling of world models and reinforcement learning creates a feedback loop that accelerates model iteration and performance [52] Group 5: Industry Trends and Future Directions - The importance of data is being redefined, with a shift towards the ability to model the world rather than just relying on raw data [56] - Companies are focusing on enhancing the "modeling capacity" of their systems, which is crucial for intelligent driving [60] - The evolution of intelligent driving systems is moving towards a stage where AI can independently understand environments and refine strategies, marking a significant advancement in the industry [62]

行 业 研 究 2026 年 02 月 01 日 投资评级：看好（维持） 行业走势图 数据来源：聚源 -10% 0% 10% 19% 29% 38% 2025-02 2025-06 2025-10 汽车 沪深300 汽车研究团队 邓健全（首席分析师） 证书编号：S0790525090003 赵悦媛（联席首席分析师） 证书编号：S0790525100003 行业重点新闻 （1）特斯拉 Robotaxi 战略揭秘：双座 Cybercab 主攻九成出行场景，Model Y/Robovan 补位；（2）特斯拉首次披露 FSD 付费用户数据：约 110 万人，占公 司累计车辆销量约 12%；（3）上海推出汽车置换更新补贴：购买新能源车补贴 车价 8%，最高不超 1.5 万元；（4）国家能源局：2027 年底将建成 2800 万个充 电设施，预计拉动投资 2000 亿元以上；（5）文远知行发布通用仿真模型 WeRide GENESIS，几分钟即可构建仿真城市环境；（6）玉柴发布全球首个飞轮增程技 术品牌，最高发电效率突破 4.8kWh/L；（7）潍柴超 10 万台份额超 50%，龙擎/ 玉柴领涨，燃气重卡动力 2025 ...

汽车驾驶辅助系统领域首个强制性国家标准发布 据工业和信息化部，强制性国家标准《轻型汽车自动紧急制动系统技术要求及试验方法》（GB 39901 —2025），将于2028年1月1日正式实施。 岚图汽车以介绍方式在香港联合交易所上市获中国证监会备案 1月28日，中国证监会国际合作司发布关于岚图汽车科技股份有限公司境外发行上市备案通知书，公司 以介绍方式在香港联合交易所上市。 蔚来世界模型NWM新版本正式推送，新增城区领航换电功能 1月28日，蔚来宣布"蔚来世界模型NWM"新版本正式开启推送，首批将为超过46万辆"Banyan榕"车型推 送。新版本新增城区领航换电功能，可支持全国超过2000座二代及以上城区换电站。 雷军：小米SU7 Ultra将正式入驻GT7 1月28日，雷军发文称，北京时间1月29日14:00，小米SU7 Ultra将正式入驻《Gran Turismo 7/跑车浪漫旅 7》。这是首台在GT游戏里上线的中国品牌车型。 东风纳米06海外版车型在哥斯达黎加上市 1月28日，据东风汽车消息，东风VIGO（纳米06海外版）上市活动暨大客户签约仪式于近日在哥斯达黎 加首都圣何塞东风展厅举行，现场获270辆终 ...

点击输入图片描述（最多30字） 点击输入图片描述（最多30字） 【鸣鸣很忙上市大涨88%】中国最大的休闲食品饮料连锁零售商——湖南鸣鸣很忙商业连锁股份有限公司在港交所主板挂 牌上市，开盘大涨88.08%，截至收盘，涨幅回落至69.06%，报400港元/股，总市值达862.04亿港元。鸣鸣很忙此次全球发 售1410.1万股H股，公开发售阶段认购倍数超过1500倍，8家基石投资者合计认购15亿港元，包括腾讯、淡马锡与贝莱德 等。公司预计净筹资30亿港元以上，发行市值约500亿港元。鸣鸣很忙旗下拥有零食很忙、赵一鸣零食两大品牌，最新数据 显示，截至2025年11月30日，鸣鸣很忙门店数达21041家。（更多内容，点击查看《两位85后卖零食，开店两万家，身价均 超200亿港元，鸣鸣很忙刚刚IPO》）（中国基金报） 点击输入图片描述（最多30字） 【英伟达员工晒年会最牛年终奖：黄仁勋亲笔签名皮衣】近日，英伟达CEO黄仁勋陆续到访上海、北京、深圳，分别参加 了三地分公司举办的新年活动。1月27日，有网友在社交媒体晒出一件黄仁勋亲笔签名的皮衣。该网友表示，自己本来想签 在胸前，结果黄仁勋直接把他翻个身按住并称"Don't ...

Group 1 - Jack Ma emphasizes the importance of teaching children how to effectively use AI rather than debating its necessity, highlighting curiosity, imagination, and creativity as key skills for the AI era [2] - The Ma Yun Rural Teacher Program, initiated by the Jack Ma Foundation in 2015, aims to enhance rural education quality through funding and professional training, selecting 100 teachers annually for support [2] Group 2 - Hunan Mingming Henbang, China's largest snack and beverage retail chain, saw its stock surge by 88.08% upon listing on the Hong Kong Stock Exchange, closing with a market cap of HKD 86.204 billion [4] - The company issued 14.101 million H shares, with a subscription rate exceeding 1500 times, and anticipates net proceeds of over HKD 3 billion from the offering [4] Group 3 - Tesla plans to cease production of Model S and Model X vehicles to transition towards a fully autonomous future, utilizing the "Optimus" robot factory for production [8] - Elon Musk stated that this change is part of Tesla's overall shift towards automation, although support for existing Model S and Model X vehicles will continue [8] Group 4 - SoftBank is reportedly negotiating to invest up to $30 billion in OpenAI, which is seeking to raise up to $100 billion in new funding, potentially valuing the company at $830 billion [10] - OpenAI's largest shareholder, SoftBank, previously invested $22.5 billion in December, increasing its stake to approximately 11% [10] Group 5 - Meta reported Q4 revenue of $59.89 billion, a 24% year-over-year increase, with advertising revenue also rising by 24% to $58.14 billion [11] - The company expects Q1 sales to range between $53.5 billion and $56.5 billion, with a post-earnings stock increase of over 4% [11] Group 6 - Ideal Auto is restructuring its R&D framework, splitting the autonomous driving team into three main groups, with a focus on software and hardware integration [10] - The restructuring aims to enhance the efficiency of the development process and align with the company's strategic goals [10] Group 7 - Amazon announced a reduction of nearly 16,000 jobs as part of an organizational restructuring while continuing to invest in strategic areas [17] - The company aims to balance workforce adjustments with ongoing recruitment in critical functions [17] Group 8 - Longting Technology completed a new round of financing totaling 500 million yuan, with investors including national and local AI industry funds [18] - The funds will be used to accelerate product development and enhance the company's market position [18]

Group 1 - WeRide announced the launch of its self-developed general simulation model WeRide GENESIS, which bridges Physical AI and Generative AI, accelerating the development, training, and commercialization of autonomous vehicles [2] - Lingbo Technology, a subsidiary of Ant Group, released the embodied large model LingBot-VLA, which is fully open-sourced, providing model weights and a complete codebase for data processing, efficient fine-tuning, and automated evaluation [2] - OpenAI introduced Prism, a research tool powered by GPT-5.2, allowing researchers to write papers and collaborate on research, available for free to all ChatGPT personal account users [2] - Xiaomi, in collaboration with Ant Group, launched a parking payment feature on its smart glasses using the GPASS technology framework, enabling users to pay for parking without using their phones [2]

文远知行-W(00800)尾盘涨近3%，截至发稿，涨2.81%，报23.44港元，成交额1583.07万港元。 据介绍，依托生成式AI技术，WeRide GENESIS能够在虚拟世界中快速"生成世界"——只需几分钟即可 构建高度真实的仿真城市环境，并精准复刻现实道路中难以高频获取的极端长尾场景，使自动驾驶系统 得以在虚拟环境中完成高强度训练与验证。同时，通过不断将仿真生成的经验反哺真实道路运行， WeRide GENESIS显著提升了自动驾驶算法应对真实物理世界复杂场景的能力与迭代效率，并大幅降低 传统道路测试所需的时间与成本。 消息面上，据文远知行-W官微消息，1月28日，公司正式发布自研通用仿真模型——WeRide GENESIS，构建起物理AI(Physical AI)与生成AI(Generative AI)之间的融合桥梁，贯通"现实物理世 界"与"虚拟仿真世界"，将加速自动驾驶汽车的规模化研发、训练与商业落地。 ...

据介绍，依托生成式AI技术，WeRide GENESIS能够在虚拟世界中快速"生成世界"——只需几分钟即可 构建高度真实的仿真城市环境，并精准复刻现实道路中难以高频获取的极端长尾场景，使自动驾驶系统 得以在虚拟环境中完成高强度训练与验证。同时，通过不断将仿真生成的经验反哺真实道路运行， WeRide GENESIS显著提升了自动驾驶算法应对真实物理世界复杂场景的能力与迭代效率，并大幅降低 传统道路测试所需的时间与成本。 消息面上，据文远知行-W官微消息，1月28日，公司正式发布自研通用仿真模型——WeRide GENESIS，构建起物理AI(Physical AI)与生成AI(Generative AI)之间的融合桥梁，贯通"现实物理世 界"与"虚拟仿真世界"，将加速自动驾驶汽车的规模化研发、训练与商业落地。 智通财经APP获悉，文远知行-W(00800)尾盘涨近3%，截至发稿，涨2.81%，报23.44港元，成交额 1583.07万港元。 ...