自动驾驶之心团队推出的教研一体轻量级解决方案，支持感知、定位、融合、导航、规划等多个功能平台，阿克曼底盘。 重磅！预售来啦。面向科研&教学级自动驾驶全栈小车黑武士系列001正式开售了。世界太枯燥了，和我们一起做点有意思的事情吧。 原价36999元，现 在下单赠送3门课程（ 模型部署+点云3D检测+多传感器融合 ），优先锁定的安排组装发货。 这两个月订单排满了，正在不断组装调试，5台及以上订单可以优惠哦！欢迎高校和研究院所批量采购。感兴趣的同学可以早点下单哦～ 1）黑武士001 黑武士支持二次开发和改装，预留了众多安装位置和接口，可以加装相机、毫米波雷达等传感器； 2）效果展示 我们测试了室内、室外、地库等场景下感知、定位、融合、导航规划等功能； 整体功能介绍 本科生学习进阶+比赛；√ 研究生科研+发论文；√ 研究生找工作+项目；√ 高校实验室教具；√ 培训公司/职业院校教具；√ 户外公园行驶 点云3D目标检测 室内地库2D激光建图 室内地库3D激光建图 上下坡测试 室外大场景3D建图 室外夜间行驶 6）软件说明 软件与语言框架：ROS、C++、python 支持一键启动，提供开发环境 3）硬件说明 | 主要传 ...

Core Insights - The article discusses the rapid advancements in end-to-end autonomous driving, particularly focusing on Vision-Language-Action (VLA) models and their applications in the industry [2][3]. Group 1: VLA Model Developments - The introduction of AutoVLA, a new VLA model that integrates reasoning and action generation for end-to-end autonomous driving, shows promising results in semantic reasoning and trajectory planning [3][4]. - ReCogDrive, another VLA model, addresses performance issues in rare and long-tail scenarios by utilizing a three-stage training framework that combines visual language models with diffusion planners [7][9]. - Impromptu VLA introduces a dataset aimed at improving VLA models' performance in unstructured extreme conditions, demonstrating significant performance improvements in established benchmarks [14][24]. Group 2: Experimental Results - AutoVLA achieved competitive performance metrics in various scenarios, with the best-of-N method reaching a PDMS score of 92.12, indicating its effectiveness in planning and execution [5]. - ReCogDrive set a new state-of-the-art PDMS score of 89.6 on the NAVSIM benchmark, showcasing its robustness and safety in driving trajectories [9][10]. - The OpenDriveVLA model demonstrated superior results in open-loop trajectory planning and driving-related question-answering tasks, outperforming previous methods on the nuScenes dataset [28][32]. Group 3: Industry Trends - The article highlights a trend among major automotive manufacturers, such as Li Auto, Xiaomi, and XPeng, to invest heavily in VLA model research and development, indicating a competitive landscape in autonomous driving technology [2][3]. - The integration of large language models (LLMs) with VLA frameworks is becoming a focal point for enhancing decision-making capabilities in autonomous vehicles, as seen in models like ORION and VLM-RL [33][39].

Investment Rating - The report maintains a "Recommended" investment rating for the industry [10]. Core Insights - The report highlights that Pony.ai is approaching a profitability inflection point driven by AI-powered autonomous driving ecosystem and accelerated global expansion [4][10]. - The company reported a total revenue of $13.979 million in Q1 2025, representing a year-on-year growth of 11.6%, primarily driven by Robotaxi services [4]. - The gross margin decreased to 16.6%, down 440 basis points year-on-year, due to changes in revenue structure [5]. - Significant operational losses were reported, with an operating loss of $56 million, an increase of 61.4% year-on-year, reflecting a temporary imbalance between technological investment and commercialization [5][10]. Revenue and Profitability - Revenue growth was robust, with Robotaxi service revenue reaching $1.73 million, a staggering increase of 200.3% year-on-year [4]. - The on-demand charging revenue surged by 800%, indicating rapid expansion of commercial operations in first-tier cities in China [4]. - Operating expenses increased by 56.3% to $58.359 million, primarily due to investments in Gen 7 production and R&D [5]. Strategic Partnerships and Technological Advancements - Pony.ai has established strategic partnerships with Tencent Cloud and Uber to enhance its commercial deployment and leverage AI capabilities [6][8]. - The company has received permission to begin L4 level Robotaxi testing in Luxembourg, marking a significant step in its technological advancement [8]. - The seventh-generation autonomous driving system was showcased at the Shanghai Auto Show, featuring a 100% automotive-grade autonomous driving kit [8]. Global Expansion and Market Presence - Pony.ai has expanded its operations to over 2,000 square kilometers across various regions, including China, Europe, East Asia, and the Middle East [9]. - The company has recorded over 45,000 kilometers of freight service since launching cross-province freight services in Beijing-Tianjin [9]. - A strategic partnership with WeChat's "Mobile Travel Service" platform has led to a 20% increase in registered users in Q1 2025, enhancing its market reach [9]. Investment Recommendations - The report suggests that investors should focus on Pony.ai's technological leadership and commercialization capabilities in the autonomous driving sector, positioning it as a core investment opportunity in the AI mobility revolution [11].

Core Viewpoint - The article highlights various companies' strategic expansions and innovations in international markets, particularly focusing on Southeast Asia and the Middle East, showcasing significant growth opportunities in these regions [3][4][5]. Group 1: Company Expansions - LoBo Fast Run plans to enter the Southeast Asian market by the end of 2025, focusing on Singapore and Malaysia for its autonomous taxi services [8]. - Transsion Holdings is reportedly exploring the two-wheeler electric vehicle market, having incubated this project for nearly three years, primarily targeting Africa and South Asia [8]. - JD Logistics has launched a new B2C express delivery brand, JoyExpress, in Saudi Arabia, enhancing local e-commerce logistics [9]. - Geely Automobile has signed a joint venture agreement with Renault to strengthen its market presence in Brazil [11]. Group 2: Market Opportunities - The Chinese cross-border e-commerce export scale has surpassed 2 trillion yuan, indicating a 16.9% year-on-year growth [15]. - The short drama market in China has seen explosive growth, with the market size reaching 50.5 billion yuan, surpassing annual box office revenue for films [18]. - China's industrial robot exports have increased by 55.4% in the first five months of the year, reflecting a growing demand for high-end and customized products [19]. Group 3: Financing and Investments - Yushun Technology has completed a C-round financing, with a valuation exceeding 10 billion yuan, indicating strong investor confidence [20]. - UniUni has raised over $70 million in D1 round financing to enhance its logistics capabilities in North America [20]. - EHang Intelligent has secured significant funding to expand its urban air mobility services in Europe and Latin America [11].

Core Insights - The article discusses the advancements in autonomous driving technology in Beijing, highlighting the city's efforts to create a comprehensive ecosystem for smart vehicles and infrastructure [1][3][5]. Group 1: Industry Developments - Beijing has established the first high-level autonomous driving demonstration zone in China, which has expanded from 60 square kilometers to 600 square kilometers [5]. - The demonstration zone has issued testing licenses for 1,098 vehicles, accumulating over 38 million kilometers in test mileage [5]. - The city is actively promoting the integration of vehicle, road, and cloud technologies to enhance the safety and efficiency of autonomous driving [3][5]. Group 2: Legislative and Safety Measures - In April, Beijing implemented the "Beijing Autonomous Driving Vehicle Regulations," which standardizes the operation of autonomous vehicles and outlines safety requirements [5]. - The National Intelligent Connected Vehicle Innovation Center has developed a comprehensive security research system to address potential risks associated with smart vehicles, including network penetration [5]. Group 3: Company Initiatives - New Stone Technology has received the first permit for autonomous delivery vehicles in the demonstration zone, marking a significant milestone in the industry [6][8]. - The company has deployed over 5,000 autonomous vehicles across 15 countries, including Japan, South Korea, Germany, and the UAE, focusing on local logistics and smart community applications [8].

近日，北京商报记者跟随"活力中国调研行"北京市主题采访团走进位于亦庄的北京市高级别自动驾驶示范区，零距离感受"聪明的 车""智慧的路""可靠的网"展现出的科技活力。"目前已实现全市600平方公里路侧基础设施智能化部署，首次在全国建成涵盖'车、路、 云、网、图'五大体系的城市级工程试验平台，助力企业自动驾驶相关技术快速迭代更新。"北京市经济和信息化局汽车处相关负责人介 绍，北京持续推进智能网联汽车国家级创新中心建设，在车联网、人工智能、大数据等关键技术和系统研发方面形成领先优势，形成上 中下游产品种类齐全的智能网联汽车产业链条。 单日分拣10万件快递 走在北京亦庄的街头，身边经常会经过一辆辆外形小巧、造型呆萌的新石器无人配送车，沿着规划的路线匀速行驶。它们没有驾驶舱， 也无方向盘，却能自主避障、识别红绿灯，精准抵达每一个投递点。这样的场景，在亦庄已成日常。 顺着这辆无人车的日常路线，不远处便是新石器无人车公司（以下简称"新石器无人车"）设立于亦庄的共配中心，这也是全球首个以无 人车为核心的智慧共配中心。北京商报记者走进去看到，一层整齐排列着形态功能各异的无人车，专为最后10公里配送场景设计的轻量 级无人配送车的 ...

Key Developments - Indonesia Discovery Program 2025: Healthcare aims to assist Chinese companies in navigating the Indonesian healthcare market through a three-day on-site visit program [2] - LoBo has received approval to expand its autonomous driving tests to Tung Chung, Hong Kong, with plans to enter Southeast Asia by the end of 2025 [3] - Transsion Holdings is reportedly exploring the two-wheeled electric vehicle market, having incubated the project for nearly three years [3] - WeRide has signed a cooperation agreement with Uber and Dubai's RTA to deploy commercial Robotaxi services in Dubai [4] - JD Logistics has launched a new B2C express delivery brand, JoyExpress, in Saudi Arabia [4] - Lao Pu Gold has opened its first store in Singapore, marking a significant step in its global expansion [5] Company Updates - JST Group Corporation has received approval from the China Securities Regulatory Commission for its IPO in Hong Kong, planning to issue up to 137,184,600 shares [6] - Beijing Yuxin Technology has submitted its listing application to the Hong Kong Stock Exchange [6] - Geely has signed a joint venture agreement with Renault to enhance its brand presence and market coverage in Brazil [6] - EHang has signed memorandums of understanding with ANRA Technologies and FAdeA to collaborate on urban air mobility services [7] - JinJing New Energy and Yiwei Lithium Energy have launched a global lithium battery recycling network platform [7] Market Trends - Meitu has achieved significant success in the European market, topping app store charts in Italy, Russia, and Turkey [8] - Alibaba International Station has reported a 42% year-on-year increase in platform orders since June, with a nearly 30% growth in GMV [10] - Alibaba Cloud is set to launch its second data center in South Korea to meet the growing demand for generative AI [10] - China's cross-border e-commerce export scale has surpassed 2 trillion yuan, with a 16.9% year-on-year increase [11] Investment and Financing - Yushu Technology has completed a C round financing, with a pre-investment valuation exceeding 10 billion yuan [16] - UniUni has raised over $70 million in D1 round financing to enhance its logistics capabilities [16] - E-ink Technology has secured approximately $123 million in E round financing for AI platform upgrades [16] - Raytheon Technology has completed a C round financing of several hundred million yuan to expand R&D and overseas markets [17]

Core Viewpoint - The article discusses the advancements in autonomous driving technology, particularly focusing on Xiaopeng Motors' recent paper presented at CVPR 2025, which validates the scaling laws in the context of autonomous driving and introduces new standards for computing power in Level 3 (L3) autonomous vehicles [4][6][22]. Group 1: Scaling Laws and Model Performance - Xiaopeng Motors' paper systematically verifies the effectiveness of scaling laws in autonomous driving, indicating that larger model parameters lead to improved performance [4][6]. - The research establishes a clear power-law relationship between model performance, parameter scale, data scale, and computational power, originally proposed by OpenAI [4][6]. Group 2: Computing Power Standards - The paper introduces a new computing power standard of 2000 TOPS for L3 autonomous driving, highlighting the exponential increase in computational requirements as the driving level advances [8][20]. - For L2 systems, the required computing power ranges from 80 to 300 TOPS, while L3 systems necessitate thousands of TOPS due to the complexity of urban driving scenarios [8][20]. Group 3: VLA and VLM Model Architecture - Xiaopeng's VLA (Vision-Language-Action) model architecture integrates visual understanding, reasoning, and action generation capabilities, requiring substantial computational resources [10][12]. - The architecture's visual processing module alone demands hundreds of TOPS for real-time data fusion from multiple sensors [10][12]. Group 4: Comparison of Onboard and Data Center Computing Power - The article differentiates between onboard computing power, which focuses on real-time data processing for driving decisions, and data center computing power, which is used for offline training and model optimization [12][15]. - Onboard systems must balance real-time performance and power consumption, while data centers can leverage significantly higher computational capabilities for complex model training [12][15]. Group 5: Market Dynamics and Competitive Landscape - The market for AI chips in autonomous driving is dominated by a few key players, with NVIDIA holding a 36% market share, followed by Tesla and Huawei [20]. - The competitive landscape has shifted significantly since 2020, impacting the development of AI chips and their applications in autonomous driving [17][20].

Core Viewpoint - The article discusses the ongoing evolution of automotive intelligence, focusing on data flow from acquisition to application, and highlights key developments in the industry, including Tesla's upcoming Robotaxi trial and advancements in AI technology by various automakers [1][2][3]. Industry News - Tesla will launch Robotaxi trial operations in Texas on June 22, 2025, with an initial fleet of 10 Model Y vehicles [14]. - Xiaopeng G7 is now available for pre-order at a price of 235,800 yuan, featuring the Turing AI chip and AR-HUD technology [17]. - GAC Aion's autonomous vehicles equipped with the seventh-generation autonomous driving system are undergoing public road tests in Guangzhou and Shenzhen [5]. - WeRide collaborates with Uber and Dubai's Road Traffic Authority to deploy commercial Robotaxi services in Dubai [26]. - Black Sesame Intelligence showcases its advanced chip technology at the 2025 Hong Kong Auto Show, emphasizing its role in the evolution of new automotive technologies [21]. High-Frequency Core Data Updates - The penetration rate of 800万 pixel cameras in passenger vehicles reached 32.9% in April 2025, up 18.2% year-on-year [6]. - NVIDIA's driving chip penetration in passenger vehicles increased to 24% in April 2025, with Horizon, Tesla FSD, and Huawei chips holding 13%, 7%, and 9% market shares respectively [6]. - The penetration rate of L2 and above autonomous driving functions in passenger vehicles was 24.3% in April 2025, reflecting an 11% year-on-year increase [8]. Smart Driving - The penetration rate of L2 and above functions in passenger vehicles increased by 4.3% year-on-year in April 2025 [7]. - The penetration rates for front-view cameras, forward millimeter-wave radars, and laser radars were 65.8%, 55.8%, and 8.2% respectively in April 2025 [8]. Smart Cockpit - The penetration rates for 10-inch and above central control screens, 10-inch and above LCD instruments, HUDs, and cockpit domain controllers were 85.6%, 51.4%, 18.2%, and 32.2% respectively in April 2025 [8]. Smart Connectivity - The penetration rates for OTA and T-BOX in passenger vehicles were 76.3% and 62.7% respectively in April 2025 [10].

这种被称为"远程操作"的技术方案，在自动驾驶行业已形成两种典型实践模式。以谷歌(GOOGL.US)旗 下Waymo为代表的渐进式路线，采用"车辆自主决策为主，人工远程介入为辅"的混合模式。当自动驾驶 系统遭遇道路施工、紧急车辆通行等复杂场景时，车辆会通过蜂窝网络向远程安全员发起咨询，由人类 判断是否放行。这种模式在Waymo的运营中已形成标准化流程，其前CEO约翰·克拉夫西克曾强调"软件 始终掌握最终决策权"，人类仅在必要时提供辅助判断。 与之形成对比的是其他企业采用的"虚拟驾驶员"方案，由远程操作员直接通过5G网络操控车辆方向 盘、油门和制动系统。这种模式对网络稳定性要求极高，卡内基梅隆大学自动驾驶专家菲利普·库普曼 教授指出："在公共道路场景下，蜂窝网络延迟和连接中断风险始终存在，当前技术条件下远程驾驶本 质上仍属不可靠方案。"他特别警示，当测试车辆规模从10辆级扩大至百万辆级时，通信中断概率将呈 指数级上升。 特斯拉的自动驾驶出租车方案呈现独特的技术路径特征。马斯克今年1月曾承诺将推出基于"完全无人监 督"系统的付费服务，但5月采访时改口称，初期运营将限定在奥斯汀特定安全区域，避开复杂路口，并 配置远程 ...