Workflow
四维图新发布智能驾驶技术新布局 深化产业协同生态构建

Core Insights - The company has unveiled its latest strategic developments in the intelligent driving sector during the 2023 Shanghai International Auto Show, showcasing its full-stack technical capabilities in high-precision mapping, vehicle-mounted chips, and advanced driving assistance solutions [1] - The company has established deep strategic partnerships with technology firms such as Qualcomm and Volcano Engine [1] Technical Architecture - The company launched two core product lines based on the Snapdragon® Digital Chassis™ solution, optimizing computing power utilization by 27% and supporting urban NOA Lite functionality [3] - A joint development with Zhuoyue Technology has achieved 0.1-meter-level recognition accuracy for stationary obstacles in complex road conditions through sensor hardware and map data integration [3] Safety Technology Framework - The company introduced a five-layer intelligent driving demand model derived from Maslow's hierarchy of needs, breaking down safety performance into seven core modules, with speed control accuracy of ±1.5 km/h and lateral control error within 0.3 meters [5] - The integration of dynamic high-precision map data has reduced the warning response time for temporary construction zones to 200 milliseconds [5] Industry Collaboration - The collaboration with Volcano Engine has led to breakthroughs in multi-modal interaction for cockpit systems, achieving a voice recognition accuracy of 98.6% in noisy environments and supporting real-time conversion of 12 dialects [7] - Partnership with Hangsheng Electronics has resulted in a 35% reduction in domain controller costs, with plans for mass production of over one million units by 2024 [7] Production Data - In 2023, the company added 23 automotive manufacturers covering 47 models for its advanced driving assistance systems, with basic ADAS product lines achieving large-scale deployment in models priced below 150,000 yuan [9] - The company's self-developed visual perception module has reduced costs by 40% compared to industry averages, while the millimeter-wave radar solution achieves a detection range of 150 meters with an angular resolution of ±0.2 degrees [9] Technological Validation - The CEO emphasized that the company's solutions utilize a white-box verification architecture, with all algorithm modules certified to ISO 26262 ASIL-B level, achieving a 99.3% success rate in recognizing construction scenarios during 100,000 kilometers of road testing [9][10] Integrated Urban Trials - The company has disclosed participation in integrated vehicle-road-cloud trials in six cities, including Beijing and Shanghai, with roadside equipment supporting L4-level autonomous driving capabilities [12] - The V2X technology implemented has reduced traffic event warning information transmission latency to within 50 milliseconds, resulting in a 63% reduction in accident rates compared to purely visual solutions [12] - The recent technology release signifies the establishment of a complete ecosystem in China's intelligent driving supply chain, encompassing "chips-algorithms-data-cloud services," with costs for integrated driving and parking solutions expected to drop to the 2,000 yuan range, accelerating the adoption of L2+ level autonomous driving in the passenger vehicle market [12]