40款车平均通过率不足36%，争议测试背后的自动驾驶迷局

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].