Core Viewpoint - The strategic cooperation agreement among Youjia Innovation, Dingshang Iron, and Bihu Technology aims to develop and deliver 3,000 autonomous logistics vehicles that meet automotive standards, enhancing market competitiveness and establishing a leading industry position [1][2]. Group 1: Cooperation Framework - The collaboration focuses on leveraging each party's strengths in operational management, R&D, and L4 autonomous driving technology to create a comprehensive strategic partnership [1]. - The agreement emphasizes "complementary advantages, collaborative development, and win-win creation" in the research, manufacturing, and commercialization of autonomous logistics vehicles [2]. Group 2: Responsibilities and Contributions - Youjia Innovation will ensure compliance with national and industry automotive certification standards, leading the integration of smart driving and chassis systems, and conducting comprehensive testing to deliver operational vehicles [1][2]. - Bihu Technology will lead the development of the vehicle chassis system, managing the entire process from component selection to performance validation, ensuring structural stability and operational reliability [2]. - Dingshang Iron will define the product based on deep insights into logistics scenarios, providing practical recommendations for product features, performance metrics, and market expectations [2]. Group 3: Market Deployment and Future Prospects - The parties agree to scale deployment by integrating manufacturing and operational resources, enhancing logistics efficiency and cost-effectiveness [2]. - The board believes that this cooperation will effectively reduce logistics costs and provide a replicable model for cost reduction and efficiency improvement in the industry [3]. - The long-term outlook for the L4 autonomous logistics vehicle business is positive, with significant progress and multiple orders exceeding 100 units already secured [3].

Core Viewpoint - Tesla's submission to the California Public Utilities Commission (CPUC) reveals that its "Robotaxi" service still relies on human drivers and remote operators, contradicting its claims of full autonomy [1][3][4] Group 1: Operational Details - Tesla's ride-hailing service requires a human driver in the vehicle and a remote operator, highlighting a significant operational gap compared to Waymo's fully autonomous service [3] - The service operates under the Transport Charter Program (TCP) with a Level 2 (L2) driver assistance system, necessitating a licensed human driver to monitor and take control when needed [3][4] - Tesla has deployed local remote operators in Austin and the Bay Area, who must pass background checks and training, providing an additional layer of safety [3][4] Group 2: Marketing and Regulatory Issues - Tesla argues against mandatory passenger consent for L2 technology use, claiming that user agreement during app registration suffices [5][7] - The company opposes Waymo's proposal to restrict the use of terms like "autonomous" or "Robotaxi" for L2 services, asserting that existing advertising laws are adequate [6][7] - Tesla's contradictory stance of claiming its vehicles are not autonomous while simultaneously marketing them as "Robotaxi" raises questions about its compliance with advertising regulations [6] Group 3: Safety Claims and Data Reporting - Tesla cites its Vehicle Safety Report, claiming that the probability of accidents decreases to 1/7 when using the Full Self-Driving (FSD) system [9] - The self-reported safety data has faced criticism due to Tesla's lack of comprehensive takeover data, which is required for autonomous vehicle licensing [9]

研究团队通过招募专家与新手司机完成危险检测、可用性识别、异常检测三类任务，结合眼动数据划分 注意力阶段，再将不同阶段注意力融入AxANet、UniAD等专业算法及DriveLM等视觉语言模型 （VLM），最终揭示：人类与算法注意力的核心差异并非"空间定位"，而是"语义理解"——人类能通过 自上而下的认知赋予场景特征语义优先级，而算法难以自主习得这一能力。该发现为自动驾驶算法的性 能提升提供了非规模化的新路径，对资源受限的车载实时系统部署具有重要实践意义。（青山） 【环球网科技综合报道】2月22日消息，据清华大学智能产业研究院（AIR）消息，一篇发表于《npj Artificial Intelligence》2026年2月的研究，以自动驾驶这一安全关键领域为载体，首次通过"人类眼动追 踪实验+算法对比验证"的双轨设计，系统性拆解了人类与算法视觉注意力的本质差异。 其核心价值在于提出人类驾驶注意力的三阶段量化划分框架，并证实：算法视觉理解的核心缺陷是缺 乏"语义显著性提取能力"，而融入人类检查阶段的语义注意力，能以经济高效的方式填补专业算法 的"语义鸿沟"与大模型的"接地鸿沟"，无需依赖大规模预训练。 来源：环球 ...

九识首席执行官孔旗表示，这一合作不仅带来车队规模增长，更重要的是推动公司技术路线与商业模式 发生结构性变化。 （原标题：运营成本年降六成，九识押注"无图驾驶"） 在自动驾驶物流赛道仍普遍围绕盈利模型反复试探之际，一家选择"先规模、后利润"路径的企业，正试 图借助平台资源完成跨越式跃迁。 中国无人配送车企业九识智能（Zelos Technology，以下简称"九识"）近日披露，在与阿里巴巴旗下物 流平台菜鸟的自动驾驶车辆业务完成整合后，公司将目标锁定为打造规模达20亿美元的业务体系，并同 步推进核心技术投入与海外扩张。 押注无图，规模优先 孔旗透露，九识计划在未来三至五个月推出新一代自动驾驶系统，使无人配送车具备无需高精地图即可 运行的能力。该技术若顺利落地，将使车辆能够进入此前难以覆盖的区域，包括社区道路与乡村场景， 从而扩大服务半径并降低部署成本。 公司披露数据显示，2025年下半年整体运营成本下降约60%，远程人工接管费用已降至月度运营成本的 3%以下。随着车辆规模扩大与算法持续优化，其单车运营与管理成本正被进一步摊薄。 在技术投入方向上，九识表示将持续加码人工智能能力建设，包括大模型、算力平台、芯片与运 ...

Group 1 - The Livermore Chinese Concept Stock Index rose by 0.16% on February 19, with significant divergence in individual stock performances [1] - Hilltop Metal Mining and Canadian Solar both recorded gains exceeding 4%, while Futu Holdings and Amer Sports followed with gains over 2% [1] - Century Internet experienced a drop of over 9%, with other stocks like WeRide, Kingsoft Cloud, and AHG also declining by more than 3% [1] Group 2 - The Nasdaq Golden Dragon China Index fell by 0.04%, with mixed performances among popular Chinese concept stocks [1] - Notable stocks such as Pinduoduo, Trip.com, and Bilibili saw varying degrees of increase, while Li Auto and Tencent Music experienced slight declines [1] - The Livermore Chinese Concept Stock Index has shown frequent volatility recently, with a significant drop of 2.64% on February 13, where Tencent Music fell over 10% and Pinduoduo dropped over 4% [1]

当地时间2月16日，特斯拉CEO埃隆·马斯克再次重申，没有踏板和方向盘的Cybercab（赛博出租车）将 于今年4月启动生产。这已是过去半年内，马斯克第三次明确锁定同一投产时间节点。 马斯克在社交平台X上明确表示，Cybercab将是一款彻底取消方向盘和踏板的完全自动驾驶车辆。此前 外界曾猜测量产版本可能保留这两项配置作为过渡，但马斯克此番表态否定了这一可能。这款双座金色 车型承载着特斯拉向自动驾驶出行服务平台转型的战略意图，将成为其Robotaxi网络的核心运营载体。 对于投产节奏，马斯克坦承Cybercab的早期产量不会太高。他表示，Cybercab采用了大量革命性的新零 件和全新的生产工艺，产能爬坡将遵循S型曲线——"初期生产会慢得令人煎熬，但最终会快得惊人"。 特斯拉已在阿拉斯加进行高寒环境下的冬季测试，确保车辆在极端天气下的可靠性。 值得注意的是，特斯拉在兑现产品时间表方面的记录并不理想，但马斯克在半年内三次重申同一投产日 期，在其过往表态中并不多见。这款彻底颠覆传统汽车设计的车型能否如期落地，成为市场关注的焦 点。 自动驾驶已成为车企在智能化时代立足的核心竞争力，既是技术护城河的关键构成，也是商业 ...

Core Viewpoint - Tesla has launched its purpose-built autonomous taxi vehicle, Cybercab, marking a significant step in the fully autonomous driving sector, but its commercial prospects depend on regulatory approval progress [1][3]. Group 1: Vehicle Design and Features - The Cybercab is designed specifically for Tesla's emerging autonomous taxi business, differing from the 2025 Model Y, as it lacks any manual driving controls [3][4]. - This two-door vehicle is intended for complete autonomous driving, eliminating the steering wheel and pedals [3][4]. Group 2: Regulatory Challenges - Achieving legal road use for the Cybercab requires overcoming significant regulatory hurdles, as federal vehicle safety standards were established with human control systems in mind [3][6]. - Tesla may need to apply for special exemptions from regulatory agencies to meet these requirements, similar to what Zoox has done for its operations in Las Vegas and San Francisco [7][8]. Group 3: Market Position and Pricing - Currently, Tesla's autonomous taxi project primarily operates using the 2025 Model Y, with most trips still relying on human safety supervisors [4]. - Elon Musk indicated that consumers will have the option to purchase the Cybercab, which is designed for autonomous driving and is priced at approximately $25,000 [8].

Industry Overview - The global humanoid robot market is expected to see a shipment of approximately 17,800 units by 2025, representing a year-on-year growth of 508%, with a market sales value of around $440 million [1] - Chinese manufacturers dominate the market, with Zhiyuan Robotics and Yushu Technology shipping about 5,000 units combined, leading the industry [1] - The commercial application of humanoid robots is primarily focused on entertainment, education, and data collection [1] Digital Trade - China's digital service trade surplus is projected to double by 2025, reaching approximately $33 billion, driven by the expansion of cloud computing and AI sectors [2] - The surplus in telecommunications, computer, and information services is expected to be around $31.8 billion, with a year-on-year increase of nearly 30% [2] - Major Chinese tech companies like Alibaba and ByteDance are establishing robust overseas business systems in e-commerce, gaming, and social networking [2] E-commerce and Logistics - Taobao's overseas sales during the New Year festival increased by over 40% compared to the previous year, with significant growth in pet and automotive accessories [3] - JD.com has launched its self-built logistics network, JoyExpress, in Europe, covering major cities and offering same-day and next-day delivery services [7] - Cainiao has initiated Spring Festival logistics guarantees, providing rapid delivery services across over 200 cities in China [8] Automotive Industry - BYD has officially entered the Egyptian market, aiming for overseas sales of 1.3 million units by 2026, a nearly 25% increase from the previous target [6] - AITO Wenjie has partnered with Abu Dhabi Motors to enter the UAE market, marking a significant step in its global strategy [9] - The collaboration between Tencent and Uber expands Tencent's ride-hailing service to over 20 countries, enhancing user accessibility [10] Strategic Partnerships - Midea Group has signed a strategic agreement with CMA CGM to enhance cross-border logistics and technology innovation between China and the U.S. [13] - Temu has partnered with DEKRA to improve compliance and safety standards for electronic products on its platform [11][12] - Miaokelando has collaborated with SADAFCO to explore the children's cheese snack market in Saudi Arabia, leveraging local distribution channels [14]

在解决自动驾驶问题的这场高风险竞赛中，多年来已经出现了深刻的哲学和技术分歧。 一边是几乎整个汽车和科技行业，他们倡导一种叫做传感器融合的概念——一种双保险的方法，将 摄像头、雷达和激光雷达结合起来，构建一个冗余的、多层次的世界视图。 而另一边，特斯拉孤军奋战，大胆而有争议地押注于单一模式——纯粹的、基于摄像头的视觉。 特斯拉主动移除并禁用车辆上的雷达等硬件的决定遭到了广泛的质疑，但这一举措源于其对人工智 能和自然智能本质的深刻理解和根本信念。要理解特斯拉为何做出这样的抉择，首先必须了解特斯 拉究竟摒弃了什么。 什么是传感器融合？ 传感器融合的概念相当简单。它旨在利用不同类型传感器的独特优势，构建一个统一且高度鲁棒的 车辆周围环境模型。每种传感器都有其自身的优缺点，理论上，融合它们可以弥补每种传感器的不 足。 摄像头能够提供最丰富、最高分辨率的数据，像人眼一样感知世界的色彩和纹理。它们可以读取路 标上的文字，识别交通信号灯的颜色，并理解复杂的视觉环境。它们的主要缺点是容易受到恶劣天 气和弱光环境的影响，而且难以测量相对速度。 雷达在测量物体距离和速度方面表现出色，即使在恶劣天气下也能正常工作。它可以轻松穿透雨、 ...

据报道，为末端配送提供自动驾驶技术的中国无人配送车公司九识智能（Zelos Technology）已在新一 轮融资中筹集超过3亿美元。一位知情人士表示，这笔新融资是在该公司与阿里巴巴（Alibaba）旗下菜 鸟（Cainiao）的自动驾驶部门合并后数周内完成的，对公司的估值为20亿美元。今年1月底，阿里巴巴 在投资九识智能的同时，将菜鸟的自动驾驶部门与这家无人配送车公司合并。目前，九识智能同时运营 两个品牌。总部位于新加坡的九识智能还得到了蚂蚁集团（Ant Group）、Blue Lake Capital和鼎晖投资 （CDH Investments）等投资者的资金支持。去年10月，九识智能宣布完成由蚂蚁集团领投的1亿美元B4 轮融资。当时，九识智能在其声明中表示，其B轮系列融资总额已达到4亿美元。 责任编辑：王永生 据报道，为末端配送提供自动驾驶技术的中国无人配送车公司九识智能（Zelos Technology）已在新一 轮融资中筹集超过3亿美元。一位知情人士表示，这笔新融资是在该公司与阿里巴巴（Alibaba）旗下菜 鸟（Cainiao）的自动驾驶部门合并后数周内完成的，对公司的估值为20亿美元。今年1月底 ...