Core Insights - Xiaomi Auto has launched the 10 million Clips version of its end-to-end assisted driving system, now available for the Xiaomi SU7 series, including the Pro, Max, and Ultra models [1] - The upgraded assisted driving system shows significant improvements in high-speed scenarios, with enhanced stability and comfort during operation [1] - Xiaomi Auto emphasizes that assisted driving is not equivalent to autonomous driving, urging users to remain vigilant and ready to take control of the vehicle at all times [1] Summary by Categories Product Features - The 10 million Clips version offers a 57% improvement in vertical comfort compared to the 3 million Clips version [2] - Acceleration and deceleration are more linear and smooth, enhancing the overall driving experience [2] - The system has a 67% increase in successful detour rates, indicating improved flexibility in navigation [2] Performance Enhancements - Obstacle recognition has become more precise, leading to more decisive decision-making during driving [2] - The system has improved capabilities for navigating intersections, enhancing overall traffic management [2]

每经记者｜孙磊 每经编辑｜程鹏 余婷婷 7月29日晚，理想i8迎来上市，新车共推出3款车型，售价区间为32.18万至36.98万元。值得一提的是，理想i8不仅是理想汽车继MEGA后再次推出纯电产 品，也是其首款纯电SUV车型，售价与采用增程动力的理想L8（32.18万元起售）相同。 "理想i8的设计灵感来自游艇，游艇式设计可以拥有更低的风阻和能耗，更大的乘坐空间，更好的视野。"李想称，理想i8是越野车、轿车、MPV的结合 体。不过，新车当前的造型仍引得不少争议，有不少消费者认为其是一辆MPV。"还在纠结理想i8是不是MPV的朋友，建议换个视角思考一下：这是一辆 35万至40万元的'能越野的小MEGA'，不仅有MPV的空间舒适性，还有SUV的通过性和轿车的驾驶感。"李想在微博上如此表示。 图片来源：每经记者 孙磊 摄 在新车上市的同时，理想汽车美股股价也在盘前上涨，一度涨超6%。截至《每日经济新闻》记者发稿时，理想汽车美股股价由涨转跌，跌幅为3.91%。 理想汽车董事长、CEO李想表示，理想i8就是一个小版的MEGA Home，给了预算不够但喜欢MEGA的用户一个选择。 补足功课再战纯电赛道 事实上，理想i8原 ...

Core Viewpoint - The article discusses the results of a recent driving assistance system test conducted by Dongche Di, highlighting the performance of various popular models, particularly Tesla's, while raising concerns about consumer perceptions and safety standards in the autonomous driving industry [3][4][5]. Summary by Sections Test Results and Methodology - The test involved 36 models equipped with driving assistance systems, with Tesla's Model 3 and Model X achieving an 83.3% success rate in high-speed scenarios, outperforming other brands [3][4]. - The testing scenarios were based on data from three years of national traffic accident statistics, ensuring realistic conditions [3][7]. - The test included high-risk situations such as sudden obstacles and blind spots, with vehicles required to react within 1.5 seconds at speeds of 100 km/h [7]. Industry Reactions and Concerns - While Tesla's performance was highlighted, industry analysts cautioned that a single test cannot define overall safety and effectiveness [4][5]. - The increase in complaints related to driving assistance accidents, which reached 327 in 2024 (up 47% from 2023), indicates a growing consumer misunderstanding of these technologies [5]. - The article emphasizes that many automakers use "safety redundancy" as a marketing tool, potentially misleading consumers about the actual safety of these systems [5]. Testing Standards and Future Directions - The need for standardized testing in the industry is underscored, as current tests often lack comparability and transparency [8][9]. - Critics argue that the extreme scenarios used in testing do not accurately reflect real-world conditions, which could mislead consumers [8][9]. - The article notes that the China Automotive Industry Association is working on establishing a standardized testing scenario library, expected to be released by 2025 [8]. Safety and Redundancy Issues - The article discusses the "safety redundancy paradox," where the presence of multiple sensors may create a false sense of security among users, leading to dangerous behaviors such as distracted driving [10]. - It highlights that 47% of drivers become distracted when using driving assistance features, with 87% of accidents involving these systems attributed to driver inattention [10]. Regulatory Developments - There is a growing consensus on the need for quantifiable safety redundancy standards, with the EU and China moving towards regulations that require manufacturers to disclose specific performance metrics under various conditions [16][17]. - The article suggests that the focus of competition in the industry will shift from hardware quantity to the effectiveness of safety measures, emphasizing the importance of algorithm optimization over mere sensor count [19].

补足功课再战纯电赛道 事实上，理想i8原本计划在2024年发布，但由于理想MEGA在2024年的表现未达预期，理想汽车调整了纯电产品的发布节奏，将理想i8放在了今年下半年发 布，更小尺寸的理想i6则计划在今年9月发布。 每经记者｜孙磊 每经实习编辑｜余婷婷 7月29日晚，理想i8迎来上市，新车共推出3款车型，售价区间为32.18万至36.98万元。值得一提的是，理想i8不仅是理想汽车继MEGA后再次推出纯电产品， 也是其首款纯电SUV车型，售价与采用增程动力的理想L8（32.18万元起售）相同。 理想汽车董事长、CEO李想表示，理想i8就是一个小版的MEGA Home，给了预算不够但喜欢MEGA的用户一个选择。 在新车上市的同时，理想汽车美股股价也在盘前上涨，一度涨超6%。截至《每日经济新闻》记者发稿时，理想汽车美股股价涨幅为6.54%。 图片来源：每经记者 孙磊 摄 值得一提的是，2024年MEGA的市场表现未达预期后，理想汽车将"自营充电网络不足"列为关键短板，开始加速布局。李想此前在财报电话会上表示，理想 汽车计划在理想i8发布时达成2500座超充站里程碑。而截至7月28日，理想汽车的超充站数目已经超过 ...

Core Viewpoint - The article emphasizes the need for serious evaluation in the automotive industry, particularly regarding autonomous driving technologies, contrasting it with the more superficial engagement seen in fan culture [1]. Group 1: Testing and Results - A large-scale test conducted by Dongche Di evaluated over 40 models from popular brands, including Tesla, in various high-risk driving scenarios [3][5]. - None of the tested vehicles achieved zero errors in both highway and urban scenarios, with a significant performance gap observed between domestic brands and Tesla [5][10]. - Tesla's Model 3 and Model X achieved the highest scores, passing 5 out of 6 scenarios, while many domestic models failed to pass any [5][10]. Group 2: Industry Reactions - The test results sparked significant public discourse, with terms like "domestic autonomous driving collective failure" trending, yet the automotive industry largely remained silent or suggested setting national standards [10][11]. - The official endorsement of the test by entities like CCTV and the Ministry of Public Security added weight to the findings, emphasizing that current systems do not possess true "autonomous driving" capabilities [13][15]. Group 3: Testing Methodology and Concerns - The test included extreme scenarios such as "disappearing front vehicles" and "nighttime construction zone avoidance," with only 14% success in the former [17][19]. - Critics raised concerns about the variability in testing conditions, questioning the fairness of comparisons between different models [22][24]. - The article suggests that the testing methodology, while revealing industry weaknesses, may not fully represent the stability of the systems under real-world conditions [24][25]. Group 4: Future Standards and Regulations - The article discusses the potential for new industry standards, such as "failure backup capability," to be established, which could address the shortcomings revealed in the tests [26][30]. - The C-NCAP plans to introduce new evaluation criteria by 2025, focusing on sensor effectiveness and system decision-making resilience [26]. Group 5: Public Perception and Media Influence - The article highlights the emotional responses from consumers and the media's role in shaping perceptions of brands like Tesla and domestic manufacturers [27][29]. - It suggests that the competitive nature of testing and the resulting media coverage can lead to polarized opinions and misunderstandings among consumers [27][39]. Group 6: Conclusion and Recommendations - The article advocates for more rigorous and transparent testing practices in the automotive industry to foster genuine improvements and consumer trust [40][41].

Core Insights - The article discusses the launch of the Li Auto i8, a new six-seat pure electric SUV priced between 350,000 to 400,000 yuan, which aims to establish a new growth point for the company as it transitions from range-extended vehicles to pure electric models [1][2]. Group 1: Product Features and Specifications - The i8 is built on an 800V high-voltage platform and offers two battery options: 90.1 kWh and 97.8 kWh, providing CLTC ranges of 670 km and 720 km respectively, with a maximum energy consumption of 14.8 kWh per 100 km [1]. - The vehicle will feature Li Auto's self-developed Mind VLA driver assistance architecture and will be equipped with Hesai ATL LiDAR for safety redundancy [5][6]. - The i8 incorporates a silicon carbide motor, which helps maintain energy consumption between 14.6 kWh/100 km and 14.8 kWh/100 km, and offers improved space utilization by freeing up 24 mm of longitudinal space in the third row [14][15]. Group 2: Competitive Landscape - The i8 faces stiff competition in the large six-seat pure electric SUV market, particularly from models like the extended Model Y and the L90, which are priced significantly lower [3][4]. - To differentiate itself, Li Auto plans to enhance its intelligent driving capabilities and address the challenges of range and charging speed through innovations like 5C ultra-fast charging technology [4][22]. Group 3: Supplier Collaboration and Innovation - Li Auto has established a unique co-creation mechanism with suppliers, such as Hesai Technology, to ensure that user-centric design principles are integrated into product development [13][14]. - The ATL LiDAR developed for the i8 has seen significant innovations, including a 60% reduction in size and a 55% decrease in power consumption while doubling performance compared to previous models [11]. - The collaboration with suppliers extends to the development of silicon carbide modules, which enhance efficiency and reduce energy loss, contributing to improved vehicle range [19][21]. Group 4: Battery Technology and Safety - The i8 features a 5C battery from CATL and Xinwanda, capable of charging to provide a range of 500 km in just 10 minutes, with enhanced safety measures through an inverted battery pack design [22][24]. - The design of the battery pack has been optimized for space utilization, allowing for a more compact structure that increases passenger space while maintaining safety standards [24][28]. Group 5: Quality Control and Data Utilization - Li Auto emphasizes proactive quality control by integrating supplier data into its systems to identify potential issues before they arise, contrasting with the reactive approaches of other manufacturers [30][32]. - The company has developed a deep understanding of supply chain management, particularly in the context of emerging technologies like silicon carbide, to ensure high-quality production standards [31][33].

Core Viewpoint - The recent driving assistance test conducted by Dongche Di revealed that the average pass rate for nearly 40 vehicle models from over 20 brands was only 35.74%, highlighting significant safety concerns in the industry [2][3]. Group 1: Test Results and Implications - The test included 15 accident scenarios, with specific challenges in both highway and urban settings, showing that only 47% of vehicles could safely navigate construction scenarios and 58% could stop effectively for children unexpectedly crossing the road [2]. - Tesla's Model X and Model 3 outperformed other brands, successfully passing 5 out of 6 highway scenarios and 8 out of 9 urban scenarios [2]. - The results have sparked renewed discussions among car manufacturers regarding the safety of driving assistance technologies, with some executives acknowledging common technical bottlenecks in extreme scenarios [5][6]. Group 2: Industry Reactions - Various car manufacturers have expressed differing opinions on the test results, with some emphasizing the need for improved capabilities in high-speed evasion and recognition of non-standard obstacles [5][6]. - Tesla's CEO Elon Musk highlighted the company's achievement in the absence of local training data, while other companies refrained from commenting on the test [5][6]. - The test has also drawn criticism from netizens, questioning the validity of the results, which Dongche Di has defended by clarifying the testing methodology [6][7]. Group 3: Technological Challenges and Solutions - The industry faces significant challenges in enhancing vehicle perception systems, particularly in recognizing non-standard obstacles and adapting to complex driving environments [10][11]. - Solutions proposed include the integration of laser radar to improve near-field perception and the enhancement of computational power to better handle dynamic road conditions [11][12]. - The need for multi-dimensional visual fusion perception is emphasized to address the limitations of current sensor technologies, which struggle under adverse weather conditions and complex scenarios [12][13].

Core Insights - The recent testing of nearly 40 popular vehicle models' advanced driver-assistance systems (ADAS) revealed significant shortcomings, particularly in extreme scenarios, sparking widespread discussion among netizens [1][2][3] Group 1: Testing Results - The testing conducted on 36 mainstream models involved 216 collision tests across 15 high-risk scenarios, with a pass rate of only 24% on highways and 44.2% in urban settings [1][2] - In specific tests, only 47% of vehicles successfully avoided a truck in a night construction zone, and 58% failed to stop for a child crossing the street [1][2] - Over 70% of models could not avoid collisions with stationary vehicles in the "disappearing front vehicle" test, highlighting the limitations of current L2-level ADAS in extreme conditions [2][3] Group 2: Industry Response - Responses from automakers varied, with Huawei's brands avoiding direct comments while emphasizing user engagement statistics, indirectly questioning the test's representativeness [4] - Tesla's CEO Musk claimed the company achieved the highest results in the tests despite lacking local training data, while the company's vice president downplayed the importance of rankings [6][8] - Li Auto's sales manager acknowledged the test as a reflection of industry-wide technical limitations in extreme scenarios [8] Group 3: Testing Methodology Controversies - The fairness of the testing methodology has been questioned, with discrepancies in testing conditions such as speed and following distances affecting results [11][12] - Allegations of human intervention during tests have emerged, particularly regarding the question of whether the steering wheel was forcibly corrected, impacting the outcome [12] - The design of certain test scenarios has also been criticized for exceeding the safety boundaries of L2 systems, raising concerns about the validity of the results [13][16] Group 4: Regulatory and Safety Implications - The testing has prompted a broader industry reflection on safety standards, with calls for mandatory inclusion of "failure mitigation capabilities" in industry standards [10][17] - A report indicated that over 80% of ADAS-related accidents occur during the day, suggesting a psychological blind spot for users in seemingly safe environments [17] - The Ministry of Public Security has reiterated that current ADAS systems do not equate to "autonomous driving," emphasizing the driver's legal responsibilities [19]

Core Viewpoint - The article discusses the recent auxiliary driving simulation tests conducted by the automotive information platform Dongche Di, highlighting the performance of various car brands and the implications for safety and technology in the industry [1][2]. Group 1: Test Results and Industry Response - The simulation tests involved 15 scenarios using nearly 40 models from over 20 brands, revealing that only 47% of vehicles could safely navigate a construction site at night, and 58% could effectively stop for a child unexpectedly crossing the road [1]. - Tesla achieved the best results in the tests, with 15 out of 36 vehicles receiving a "0 pass" rating in the high-speed accident scenario simulation [2]. - Industry leaders, including Lantu and AITO, expressed concerns about the testing methods and the need for improved technology in recognizing non-standard obstacles [3][4]. Group 2: Safety and Regulatory Considerations - The article emphasizes that L2 level auxiliary driving is not equivalent to autonomous driving, and drivers remain the primary safety responsibility [1][5]. - The head of the Traffic Management Bureau highlighted that all current "smart driving" systems in the market are still in the auxiliary driving phase, requiring human control [5][6]. - There are calls for stricter industry standards, including mandatory "failure backup capabilities" and a focus on the "worst-case scenario pass rate" to enhance safety measures [4].

Core Insights - The recent evaluation of advanced driver-assistance systems (ADAS) by Dongche Di sparked significant discussion, revealing that the average pass rate for nearly 40 tested models was only 35.74% across 15 simulated high-risk scenarios [1][3] Group 1: Test Results and Industry Response - Dongche Di's program tested over 40 models from popular brands, including Tesla, NIO, and BYD, in various urban and highway scenarios, with a disappointing overall performance [1] - Tesla's Model X and Model 3 performed notably well, passing 5 out of 6 highway scenarios and 8 out of 9 urban scenarios, respectively [3] - Tesla CEO Elon Musk claimed that the company achieved the "highest score" in the tests despite lacking local training data, emphasizing ongoing improvements in safety standards [3][6] Group 2: Industry Perspectives - Various automotive companies, including AITO and Zhijie, declined to comment on the evaluation results, indicating a cautious approach to public relations [8] - Lantu Automotive's executive highlighted the common technical challenges faced by the industry, particularly in extreme scenarios, and stressed the need for improved reliability in ADAS technology [9][10] - The evaluation results prompted Dongche Di to clarify that the tests were not meant for ranking but to educate consumers about the limitations of ADAS and the importance of driver responsibility [12][13] Group 3: Regulatory and Safety Considerations - The Chinese authorities reiterated that current ADAS systems do not equate to fully autonomous driving, emphasizing that drivers remain the primary responsible party [13][14] - The government plans to enhance regulations surrounding ADAS, including clearer definitions of system capabilities and responsibilities for manufacturers [14][15]