Group 1 - TIER IV has been selected by the Japanese government's NEDO for a project to develop a comprehensive AI data platform and ecosystem aimed at enhancing industrial competitiveness in autonomous driving [1][2] - The initiative will focus on collecting large-scale real-world driving data and generating synthetic data using multimodal AI to address scenarios that are challenging to capture in reality [1][2] - The project aims to improve dataset capabilities and create an AI data platform that alleviates the development burden for individual companies, thereby boosting Japan's competitiveness in the autonomous driving sector [2] Group 2 - The project will involve three main areas: large-scale driving data collection, synthetic data generation using AI, and the development of an AI data infrastructure and ecosystem [2][5] - The collection of diverse driving data will utilize multiple vehicle models and enhance data construction efficiency through automated labeling [5] - The synthetic data generation will focus on creating high-quality training data that encompasses rare scenarios and various weather and time conditions using a multimodal generative AI model [5]

来源 | 机器之心 原文链接： 刷新NAVSIM SOTA，复旦引望提出Masked Diffusion端到端自动驾驶新框架 点击下方 卡片 ，关注" 自动驾驶之心 "公众号 戳我-> 领取 自动驾驶近30个 方向 学习 路线 >>自动驾驶前沿信息获取 → 自动驾驶之心知识星球 本文只做学术分享，如有侵权，联系删文 随着 VLA（Vision-Language-Action）模型的兴起，端到端自动驾驶正经历从「模块化」向「大一统」的范式转移。然而，将感知、推理与规划压缩进单一模型 后，主流的自回归（Auto-regressive）生成范式逐渐显露出局限性。现有的自回归模型强制遵循「从左到右」的时序生成逻辑，这与人类驾驶员的思维直觉存在本 质差异 —— 经验丰富的驾驶员在处理复杂路况时，往往采用「以终为始」的策略，即先确立长期的驾驶意图（如切入匝道、避让行人、靠边停靠），再反推当前 的短期操控动作。此外，基于模仿学习的模型容易陷入「平均司机」陷阱，倾向于拟合数据分布的均值，导致策略平庸化，难以在激进博弈与保守避让之间灵活切 换。 针对上述痛点， 复旦大学与引望智能联合提出了 WAM-Diff 框架 。该研究创新 ...

Core Insights - TIER IV has formed a capital and business alliance with Central Japan Railway Company (JR Central) to enhance first- and last-mile mobility solutions, improving access to railway stations and contributing to regional development [1][3][4] Group 1: Company Overview - TIER IV is a leader in open-source software for autonomous driving, known for its Autoware platform, which is the world's first open-source software for this technology [5] - The company aims to reshape the future of intelligent vehicles by providing scalable platforms and comprehensive solutions across software development, vehicle manufacturing, and service operations [5] Group 2: Industry Context - Japan faces mobility challenges due to population decline and a shortage of professional drivers, making autonomous buses and shuttles essential for improving access to railway stations [2] - The collaboration between TIER IV and JR Central is expected to leverage autonomous driving technology to enhance station accessibility and support regional communities [4]

本周早些时候，Waymo表示将对车队进行更新，使其自动驾驶出租车服务在停电期间具备更好的运行 能力。12月20日，Waymo曾在旧金山大停电期间暂停服务。那次停电导致当地数万人断电，并造成其 部分自动驾驶车辆在行驶途中停在路中间，从而加剧或直接引发了交通拥堵。 截至发稿，Waymo未就置评请求立即作出回应，也未说明其周四暂停服务是否因监管部门针对山洪预 警提出的要求。(作者/箫雨) 通知写道："由于美国国家气象局发布了山洪预警，服务暂时停止。"美国国家气象局已将整个旧金山湾 区的洪水监测延长至当地时间周五22点。 Waymo旧金山自动驾驶网约车 凤凰网科技讯北京时间12月26日，据CNBC报道，根据Alphabet旗下自动驾驶出租车公司Waymo在其网 约车应用上发布的乘客通知，为应对预期中的暴风雨，该公司在周四暂时停止了在旧金山湾区的自动驾 驶出租车服务。 ...

Acquisition Details - Mercedes-Benz completed the acquisition of a 3% stake in Chongqing Qianli Technology, valued at 1.34 billion yuan (US$191 million) [1] - The transaction involved Mercedes-Benz Digital Technology and Shanghai-listed Lifan Holdings, which transferred 135.6 million shares at 9.87 yuan per share [1] Stakeholder Changes - Prior to the transaction, Lifan Holdings held over 5% of Qianli's shares; post-acquisition, Mercedes-Benz became Qianli's fifth largest stakeholder [2] - The share transfer will not change the controlling shareholders or actual controllers of Qianli, nor will it affect the offer for acquisition [3] Strategic Intent - Mercedes-Benz is committed to maintaining its shareholding in Qianli for at least 12 months, indicating a long-term strategic interest [3] - This acquisition reflects Mercedes-Benz's efforts to strengthen ties and strategic collaborations in the Chinese automotive market, the largest in the world [3] AI Integration - Mercedes-Benz, along with Tesla and Volvo, became one of the first foreign carmakers approved to deploy AI chatbots in vehicles in China, signaling a potential acceleration in the regulatory process for AI services [4][5] - The "Mercedes-Benz virtual assistant" was registered as a generative AI service by the Cyberspace Administration of China's Beijing branch [5] Company Background - Lifan Holdings, originally a motorcycle manufacturer, has diversified into automobile components and has received investments from a private equity fund linked to Geely and the Chongqing municipal government [6] - The chairman of Lifan Holdings, Yin Qi, co-founded the AI firm Megvii, which is backed by Alibaba Group [7]

Core Insights - The autonomous vehicle (AV) industry is at a critical juncture, with significant investments from various companies aiming to dominate the market and transform transportation through robotaxis [1][7] Industry Overview - The global AV market is projected to grow from approximately $106 billion in 2021 to over $2.3 trillion by 2030, driven by advancements in technology such as better sensors and artificial intelligence [8] - The emergence of AVs is expected to enhance road safety, efficiency, convenience, mobility, and accessibility, while also reducing transportation costs and environmental impact [3][4][5][6] Company Highlights - **Baidu (BIDU)**: Operates fully driverless robotaxis in over 20 cities, including Beijing and Wuhan, with plans for international expansion. The Zacks Consensus Estimate for Baidu's 2026 earnings per share indicates a 16.8% year-over-year improvement [14] - **Alphabet (GOOGL)**: Waymo has achieved over 450,000 weekly paid rides and operates driverless vehicles in several U.S. cities. The Zacks Consensus Estimate for Alphabet's 2026 earnings per share suggests a 4.1% year-over-year improvement [16] - **Uber Technologies (UBER)**: Utilizes an asset-light strategy through partnerships to integrate AV technologies, allowing for rapid scaling of autonomous services. The Zacks Consensus Estimate for Uber's 2026 earnings per share has been revised upward by 2.6% in the past 60 days [19]

Core Viewpoint - The article discusses the integration of two advanced directions in autonomous driving: Vision-Language-Action (VLA) and World Model, highlighting their complementary nature and the trend towards their fusion for enhanced decision-making capabilities in autonomous systems [2][51]. Summary by Sections Introduction to VLA and World Model - VLA, or Vision-Language-Action, is a multimodal model that interprets visual inputs and human language to make driving decisions, aiming for natural human-vehicle interaction [8][10]. - World Model is a generative spatiotemporal neural network that simulates future scenarios based on high-dimensional sensor data, enabling vehicles to predict outcomes and make safer decisions [12][14]. Comparison of VLA and World Model - VLA focuses on human interaction and interpretable end-to-end autonomous driving, while World Model emphasizes future state prediction and simulation for planning [15]. - The input for VLA includes sensor data and explicit language commands, whereas World Model relies on sequential sensor data and vehicle state [13][15]. - VLA outputs direct action control signals, while World Model provides future scene states without direct driving actions [15]. Integration and Future Directions - Both technologies share a common background in addressing the limitations of traditional modular systems and aim to enhance autonomous systems' cognitive and decision-making abilities [16][17]. - The ultimate goal for both is to enable machines to understand environments and make robust plans, with a focus on addressing corner cases in driving scenarios [18][19]. - The article suggests that the future of autonomous driving may lie in the deep integration of VLA and World Model, creating a comprehensive system that combines perception, reasoning, simulation, decision-making, and explanation [51]. Examples of Integration - The article mentions several research papers that explore the fusion of VLA and World Model, such as 3D-VLA, which aims to enhance 3D perception and planning capabilities [24][26]. - Another example is WorldVLA, which combines action generation with environmental understanding, addressing the semantic and functional gaps between the two models [28][31]. - The IRL-VLA framework proposes a closed-loop reinforcement learning approach for training VLA models without heavy reliance on simulation, enhancing their practical application [34][35]. Conclusion - The article concludes that the integration of VLA and World Model is a promising direction for the next generation of autonomous driving technologies, with ongoing developments from various industry players [51].

Group 1: Company Overview - Pony AI Inc. (NASDAQ:PONY) is recognized as a leading autonomous mobility technology company, utilizing its Virtual Driver technology for large-scale production and deployment of autonomous vehicles [5] Group 2: Analyst Ratings and Perspectives - Barclays analyst Jiong Shao initiated coverage with a Hold rating and a $15 price target, expressing optimism about the future potential of robotaxis while acknowledging industry challenges [1][2] - Macquarie analyst Eugene Hsiao initiated coverage with a Buy rating and a $29 price target, highlighting 2026 as a pivotal year for the transition of robotaxis from pilot projects to commercial viability, positioning Pony AI as an industry leader [1][3] Group 3: Industry Insights - The Chinese autonomous driving industry is viewed positively, with a focus on regulatory challenges rather than technological hurdles for Pony AI [3] - Analysts suggest that Pony AI should prioritize establishing its presence in Tier-1 Chinese cities to enhance daily fares per vehicle, thereby improving unit economics [4]

Core Viewpoint - The article discusses the application of navigation information in autonomous driving, emphasizing its importance in providing lane guidance, waypoint information, and reference lines to enhance vehicle path planning and control [2][4][31]. Group 1: Navigation Information Application - Navigation information SD/SD Pro is already utilized in many production solutions, offering a rough global and local view for drivers [2]. - The core responsibilities of the navigation module include providing reference lines, which significantly reduce planning pressure by offering a predefined driving path [4]. - Additional functionalities include providing planning constraints and priorities, as well as path monitoring and replanning [5]. Group 2: Path Planning and Behavior Guidance - Global path planning at the lane level involves searching for the optimal lane sequence to reach the target lane [6]. - Behavior planning is enhanced by providing clear semantic guidance, allowing vehicles to prepare for lane changes, deceleration, and yielding in advance [6]. Group 3: Course Overview - The course titled "End-to-End Practical Class for Mass Production" focuses on practical applications in autonomous driving, covering topics from one-stage and two-stage frameworks to trajectory optimization and production experience sharing [23]. - The curriculum includes chapters on end-to-end task overview, two-stage and one-stage algorithms, navigation information applications, reinforcement learning in autonomous driving, trajectory output optimization, fallback solutions, and mass production experience [28][30][31][32][33][34][35]. Group 4: Target Audience and Course Details - The course is aimed at advanced learners with a background in autonomous driving algorithms, reinforcement learning, and programming [36][38]. - The course will commence on November 30, with a duration of three months, featuring offline video teaching and online Q&A sessions [36][39].

London is set to become the next testing ground for global autonomous driving firms after Chinese tech group Baidu confirmed plans to introduce robotaxis in the UK capital from next year. The move, an... ...