Core Insights - General Motors (GM) has decided to recall approximately 1,000 former employees of its subsidiary Cruise to refocus on passenger vehicle autonomous driving technology, aiming for levels L3 to L5 [2][3][10] - This decision comes just six months after Cruise laid off over 50% of its workforce, amounting to more than 1,000 employees, as part of a strategic shift away from autonomous taxi services [3][10] - The new strategy emphasizes safety redundancy, cost constraints, and production rhythm, led by GM's new Chief Product Officer, Sterling Anderson, who has a background in autonomous vehicle technology [4][10] Company Background - Cruise was founded in 2013 and initially targeted the consumer market with modified vehicle kits for autonomous driving, but shifted focus to providing autonomous driving software for automotive companies by 2014 [6][7] - GM acquired a 90% stake in Cruise for $581 million in 2016, allowing Cruise to operate independently while focusing on autonomous taxi development [5][6] - Despite significant investments totaling around $10 billion since GM's acquisition, Cruise has faced challenges, including a traffic accident in 2023 that led to the suspension of its autonomous testing in California [5][8][9] Financial Performance - From 2017 to 2023, Cruise has accumulated losses exceeding $8 billion, with increasing loss rates [8][9] - In June 2024, GM injected $850 million into Cruise to support its operations, indicating continued belief in Cruise's potential [9] Future Outlook - The shift back to passenger vehicle autonomous driving is seen as a strategic move to prioritize deliverable and sustainable technology, allowing for the accumulation of data and experience necessary for higher levels of automation [10][11] - The decision to abandon the autonomous taxi development reflects broader industry challenges, including high testing costs and regulatory scrutiny, which have made this path less viable in the short term [10][11]

Core Insights - The joint testing by Dongchedi and CCTV revealed that Tesla's advanced driver-assistance systems (ADAS) rank in the "fourth tier," while Chinese automakers are leading in this technology [1][10] - The results of the tests have sparked significant attention and controversy, challenging previous claims made by various car manufacturers regarding their ADAS capabilities [4][10] Testing Overview - The testing involved real-world simulations on highways and urban roads, covering 15 subjects with nearly 40 vehicle models from over 20 brands, including high-risk scenarios [1][10] - A total of 216 collision simulations were conducted to validate the performance of the ADAS across different models [1] Performance Results - Tesla's 2023 Model 3 and Model X achieved a passing rate of 83.3% in highway accident simulations, outperforming competitors like Huawei's models, which did not exceed a 50% passing rate [2] - In urban accident simulations, the Model X excelled with an 88.9% passing rate, while other brands struggled to match its performance [2] Industry Reactions - Tesla CEO Elon Musk acknowledged the testing results and emphasized Tesla's commitment to data compliance in training their systems [1] - Chinese brands such as Huawei's Zhijie and Wanjie publicly dismissed the testing results, indicating a defensive stance against the findings [1] Historical Context - Previous claims by Huawei's executives suggested their ADAS was significantly superior to Tesla's, with assertions of higher obstacle recognition capabilities [6] - Similar overstatements were made by other brands, including Xiaomi and NIO, regarding their ADAS performance, which have been contradicted by recent incidents and the testing results [8][9] Implications for the Industry - The testing results serve as a corrective measure against the exaggerated claims made by car manufacturers regarding their ADAS technologies [10] - The lack of standardized definitions for "high-level" driving assistance has led to confusion and misrepresentation in the industry, as seen with various brands coining terms like "Level 0 tier" and "high-level driving assistance" [10]

Core Viewpoint - The article discusses the recent controversy surrounding the performance of autonomous driving systems in various vehicles, particularly those from Hongmeng Zhixing, following a testing event by Dongche Di. The results highlighted significant shortcomings in the vehicles' ability to handle complex driving scenarios, prompting responses from various industry leaders. Group 1: Testing Results - Dongche Di conducted a large-scale test of 36 vehicles equipped with autonomous driving systems, including four models from Hongmeng Zhixing: Zhijie R7, Wenjie M8, Wenjie M9, and Wenjie M7. The results showed that these models had unsatisfactory pass rates in 15 different scenarios [2]. - In specific scenarios, only Wenjie M9 and Zhijie R7 successfully avoided obstacles when encountering an accident vehicle on the highway. Wenjie M9 failed to avoid obstacles in a construction scenario, while Wenjie M8 could not pass a temporary construction situation [2]. Group 2: Industry Reactions - The testing results sparked significant debate, with various automotive executives commenting. Tesla's CEO Elon Musk stated that Tesla achieved the highest results in China without local training data due to legal restrictions on data export. He emphasized that Tesla is working on adding training data to improve performance [3]. - Tesla's Vice President Tao Lin remarked that Tesla does not focus on rankings, as any test or ranking is relative and temporary, but the demand for safety is limitless [3]. - Lantu Automobile's CBO Shao Mingfeng, although not part of the test, suggested that the closed testing by Dongche Di reflects common technical bottlenecks in the industry, such as the need for improved capabilities in high-speed avoidance and recognition of non-standard obstacles. He proposed that "failure fallback capability" should be included in mandatory industry standards [4].