Summary of L3 Autonomous Driving Industry Conference Call Industry Overview - The approval of L3 autonomous driving trial runs marks readiness in regulations and the industry chain for large-scale commercialization, indicating that regulatory bodies are preparing for further promotion of autonomous driving [1][2] Core Insights and Arguments - L3 technology is initially launched in a B2B model, focusing on professional controlled testing on designated roads, which is more advantageous than directly targeting consumers due to lower education costs [1][3] - The L3 technology is expected to drive upgrades in the automotive supply chain, particularly in sensors (especially LiDAR), controllers (chip computing power), algorithms (full-stack self-research), and testing phases [1][4] - Steer-by-wire technology, as a key enabler for L3, is anticipated to improve driving experiences significantly, leading to an increase in its application ratio, benefiting related companies [1][5] - The increase in testing task requirements for L3 will provide more development opportunities for testing companies, potentially enhancing their business volume and profitability [1][5] Companies Mentioned - Changan Automobile, as one of the first approved manufacturers, is expected to benefit significantly in vehicle manufacturing [1][6] - Aptiv, a leader in steer-by-wire technology, is also viewed positively due to a high increase in Average Selling Price (ASP) [1][6] - China Automotive Research, as a key provider in the testing phase, is projected to benefit from increased task volume, leading to performance growth [1][6] Additional Important Points - The approval of L3 commercial trial runs includes two models: Changan Deep Blue S203 and BAIC Jihoo, which will be tested in Chongqing and Beijing, covering over 200 kilometers of roads, including complex urban elevated roads [2][3] - The regulatory approval signifies that the necessary preparations in regulations and the industry chain are in place for the commercial promotion of L3 autonomous driving technology [2]

Core Viewpoint - Changan Automobile has received approval for its Deep Blue automotive products to feature L3-level autonomous driving capabilities, allowing for automated driving in congested traffic conditions at speeds up to 50 km/h [1] Group 1: L3 Autonomous Driving Approval - The approval for L3-level autonomous driving was granted on December 15 [1] - The technology enables automated driving in single lanes on highways and urban expressways under traffic congestion [1] Group 2: Future Plans and Market Strategy - Changan plans to conduct pilot applications for road use in the first quarter of 2026 in collaboration with Changan Car Union Technology [1] - The company will begin product approvals and market launches for end-users once relevant national regulations and policies are officially implemented [1] - A comprehensive L3 planning strategy is in place, with subsequent models from Avita, Deep Blue, and Qiyuan expected to be introduced [1] Group 3: Market Expansion Potential - The scale of L3 technology is anticipated to expand, potentially reaching models priced around 150,000 yuan [1]

Group 1 - Hesai Technology (2525.HK) has seen a strong performance, with a stock price increase of over 32% in three consecutive trading days, significantly outperforming the Hang Seng Index which rose by 3.6% during the same period [1] - The company launched the Fermi C500, a high-performance smart control chip for LiDAR based on RISC-V architecture, and introduced the world's only photon isolation safety technology along with the upgraded ATX 256-line safety LiDAR [1] - The flagship ATX 256-line LiDAR, based on the self-developed chip platform, has a maximum detection range of 230 meters and has received over 4 million orders, with mass production planned for April 2026 [1] Group 2 - Huaxi Securities initiated coverage on Hesai with a "Buy" rating, highlighting its position as a global leader in LiDAR technology, balancing performance, quality, and cost through its self-developed ASIC technology [1] - The company has secured contracts with 22 OEMs for 120 vehicle models in the ADAS sector, ranking first in production volume, and has shipped over 20,000 units in the robotics market in a single month [1] - CICC also initiated coverage on Hesai with an "Outperform Industry" rating and a target price of HKD 182.5, noting that the company has been the top global LiDAR revenue generator from 2021 to 2024, and is expected to achieve quarterly profitability for the first time in Q4 2024 [2] - The company is projected to ship over 3 million and 10 million vehicle-mounted LiDAR units in China by 2025 and 2028, respectively, with an average annual growth rate of 58% [2]

Core Insights - Huawei's intelligent driving solution, QianKun, has become a trusted brand among users, achieving significant milestones in assisted driving and parking capabilities [2][3] - The automotive industry is transitioning towards an era of embodied intelligence, characterized by three waves: electrification, intelligence, and automation [3] Group 1: QianKun Achievements - QianKun has accumulated a total assisted driving mileage of 5 billion kilometers and has performed parking assistance 280 million times, preventing 2.71 million potential collisions [2] - The average monthly assisted driving mileage per user is 727 kilometers, with over 50% of urban driving time and an average of more than 36 smart parking instances per month [2] - The QianKun system is now installed in over 1 million vehicles, with 28 cooperative models across various brands, including AITO, Deep Blue, and Audi [2] Group 2: Industry Trends - The first wave of the automotive industry is electrification (2015-2022), which significantly reduces travel costs by replacing fuel with electricity [3] - The second wave is the intelligence wave (2022-2027), where algorithms will replace drivers, enhancing safety and user experience while reducing accident risks [3] - The third wave, starting in 2027, will focus on automation, redefining safety and experience standards, and improving transportation efficiency [3] Group 3: Technological Development Path - By 2025, Huawei aims to launch the ADS 4 system, targeting L3 technology for commercial solutions, with capabilities for L3 pilot commercial use on highways and L4 testing in urban areas [3] - In 2026, the goal is to achieve large-scale commercial use of L3 on highways and pilot commercial use of L4 in urban settings, subject to regulatory approval [3] - By 2027, L3 is expected to scale up, and L4 will enter commercial stages, paving the way for the emergence of Robotaxi and logistics solutions [3]

Core Viewpoint - The automotive industry is undergoing a transformation characterized by three waves: electrification, intelligence, and automation, leading towards embodied intelligence in vehicles [1] Group 1: Transformation Waves - The first wave (2015-2022) focuses on electrification, where electric power replaces fuel, significantly reducing travel costs [1] - The second wave (2022-2027) emphasizes intelligence, with algorithms replacing drivers, greatly enhancing safety and user experience [1] - The third wave, starting in 2027, will be centered on automation, where algorithms will redefine safety and experience standards, significantly improving travel and societal efficiency [1] Group 2: Industry Collaboration and Projections - The company QianKun Intelligent Driving has partnered with several automakers, including Seres, JAC, BAIC, SAIC, GAC, Changan, and Dongfeng, to launch 28 vehicle models [1] - The deployment of QianKun Intelligent Driving has surpassed 1 million units, indicating strong market acceptance and growth potential [1] - With the accelerated adoption of NCA in highways and urban areas, Level 3 automation is expected to scale significantly by 2027 [1]

Core Viewpoint - SAIC-GM has signed a strategic cooperation agreement with Momenta to deepen collaboration in the field of assisted driving, marking a significant step in adopting local technology solutions for intelligent driving in their vehicles [3][4] Group 1: Partnership Details - The first model to feature Momenta's R6 Flywheel model will be the Buick high-end electric sub-brand "Zhijing" L7, which will showcase advanced assisted driving capabilities [3][4] - Momenta has become a prominent player in the domestic intelligent driving supply market, with partnerships extending to various brands including Toyota and Honda, but SAIC-GM's investment in Momenta gives it a unique position [3][4] Group 2: Technology Overview - The R6 Flywheel model is based on 3 billion kilometers of real-world driving data and utilizes 70 million data points through reinforcement learning, aiming to surpass human driving capabilities [4][5] - The intelligent driving system in the Zhijing L7 includes a comprehensive sensor suite with 12 ultrasonic radars, 12 cameras, 3 millimeter-wave radars, and 1 lidar, designed with high redundancy for enhanced safety and performance [7][8] Group 3: Development History - SAIC-GM's development of intelligent driving systems has evolved through three stages, starting with the introduction of Super Cruise in North America, leading to the development of a second-generation system for the Buick GL8 [6] - The collaboration with Momenta represents an advancement in SAIC-GM's capabilities in intelligent driving, integrating Momenta's expertise in perception and algorithm models with SAIC-GM's experience in vehicle control [6]

Regulatory Landscape & Safety Concerns - The Ministry of Public Security's Traffic Management Bureau states that current "intelligent driving" systems on the market have not achieved "autonomous driving", and drivers are ultimately responsible [1] - "Hands-off, eyes-off" driving poses serious road safety risks and potential legal consequences, including civil compensation, administrative penalties, and criminal liability [1] - Public security agencies will further strengthen regulatory management in conjunction with relevant departments [1] Industry Perspective & Development - Huawei believes its intelligent driving is ready for L3 autonomous driving and calls for national regulations to allow users to sleep while using intelligent driving [1] - Huawei's 余承东 admitted to looking at his phone while using the intelligent driving system, resulting in a fine and deduction of 3 points [1]

Core Viewpoint - Huawei's HarmonyOS Intelligent Driving ecosystem is expanding, with the recent launch of the new "Xiangjie" car logo, indicating a deeper collaboration with BAIC Blue Valley and a clearer strategic direction for the HarmonyOS Intelligent Driving initiative [1][2]. Group 1: Company Developments - The "Xiangjie" brand, co-created by Huawei and BAIC Blue Valley, targets the administrative sedan market priced above 300,000 yuan [2]. - BAIC Blue Valley has committed 5.532 billion yuan for the development of the "Xiangjie" model, with 700 million yuan allocated from raised funds for three upgraded vehicle products [2]. - The "Xiangjie S9" model achieved sales of 4,154 units in June, leading the sales of new energy sedans priced above 300,000 yuan [3]. Group 2: Strategic Positioning - The HarmonyOS Intelligent Driving ecosystem consists of five brands, including "Xiangjie," which fills a gap in the administrative sedan market, differentiating itself from other models like "Wenjie" and "Zun Jie" [3][4]. - The five brands cover a market range from 100,000 to 1 million yuan, creating a closed loop of high-end (Zun Jie/Xiangjie), mid-range (Wen Jie/Zhi Jie), and entry-level (Shang Jie) vehicles [4]. - Huawei's automotive business has established a clear division of labor with traditional automakers, focusing on R&D, intelligent driving, and marketing, while manufacturers handle production [4].

Core Viewpoint - The entry of major players like Huawei has intensified competition in the dull trunk logistics autonomous heavy truck industry [1] Group 1: Company Developments - Manbang Group is increasing its management control over autonomous driving company Zhijia Technology, which is expected to remain in a net loss state throughout 2024, prompting further investment from Manbang [2] - Zhijia Technology, established in 2016, completed the industry's first fully unmanned driving operation test from warehouse to warehouse by the end of 2024 [2] - Zhijia Technology received the first operational license for autonomous freight vehicles in China in November 2018 and completed a global first demonstration operation on the S17 smart highway by the end of 2023 [2] Group 2: Market Challenges - The trunk logistics sector has faced significant challenges, with companies like TuSimple and Qianhua Technology encountering failures, leading to a shift in focus for some firms [5] - The "1+N" convoy model in trunk logistics requires a stable cargo supply system to reduce empty load rates and logistics costs, highlighting operational inefficiencies [5] Group 3: Future Prospects - Yingche Technology is reportedly considering an IPO in 2025, with significant backing from major investors and clients, which could position it as a leading autonomous driving technology company in the U.S. market [6] - NineSight, another player in the autonomous logistics space, is preparing for a Hong Kong IPO with an estimated valuation of $1.5 billion, showcasing a profitable business model [7] Group 4: Competitive Landscape - The entry of Huawei with its L3 commercial solution for highways poses a significant threat to existing autonomous driving companies in trunk logistics, as it directly addresses the high-speed scenario [10] - Companies like Manbang and Didi are adopting a defensive strategy in the face of potential full automation in freight logistics, emphasizing the need for strategic reserves to manage supply and demand dynamics [9]

Group 1 - The core viewpoint is that the L3 autonomous driving technology is being commercialized by multiple companies, with expectations for 2025 to be a significant year for its adoption [2][3][4] - Companies like GAC Group and Chery are making strides in L3 technology, with GAC set to launch its first L3 passenger vehicle in Q4 of this year [3][4] - There is a growing concern among consumers regarding the effectiveness and reliability of L2 and L3 systems, especially after incidents like the Xiaomi SU7 accident [2][6] Group 2 - The L3 technology requires advanced capabilities such as "all-weather perception" and decision redundancy, with a computing power exceeding 200 TOPS [7] - Current regulations in China do not allow L3 vehicles to operate freely on public roads, with specific conditions set for their use [6][7] - The ethical implications of L3 driving, particularly in accident scenarios, are under discussion, with a need for clear definitions of responsibility between manufacturers and drivers [9][10] Group 3 - The competition in the L3 space is intensifying as more manufacturers enter the market, indicating a shift from L2 to L3 technologies [3][4] - Industry experts emphasize the importance of adhering to regulatory standards to avoid misleading claims about L3 capabilities [6][7] - The development of L3 technology is seen as a critical factor in reshaping the automotive industry's value chain and business models [4][9]