2024H2城市NOA进入密集落地期，端到端大模型智驾方案进一步普及。2025年2月，比亚迪召开智驾发布会，天神之眼智能驾驶辅助系统下探至7万元级市场，进一步推动智驾平权。

改革开放初期，中国依靠丰富劳动力实现工业化。1980 至 2000 年间，工程师红利助力了工业化和基础设施建设，提升了生产效率并推动了技术创新与产业升级。今天，科研红利作为新兴的经济增长动力，以高投入、高风险和高回报的特性，正在引领中国经济从“制造大国”向“智造大国”转型。 科研红利依赖于高效的创新生态系统，特别是产学研合作模式。通过科研机构、企业与政府的深度协同，知识创新和技术突破得以快速转化，推动了各行业的技术进步和产业升级。产学研合作不仅能够优化资源配置，还能全面提升创新能力，成为推动经济高质量发展的重要引擎。

客车这轮盈利能创新高吗？我们认为并不是遥不可及。1）国内没有价格战。2）寡头龙头格局。3）海外无论新能源还是油车净利率远好于国内（无需投固定资产）。4）碳酸锂成本持续下行。客车这轮市值空间怎么看？小目标是重新挑战2015-2017年上一轮行业红利的市值峰值。大目标重新打开新的天花板（见证真正意义的世界客车龙头）。

终端销量表现亮眼，需求增长保持向好态势。2024 年 1-12 月，我国新能源汽车销量 1286.5 万辆，同比增长 35%，动力电池装机量 548.4GWh，同比增长 41%。终端需求保持增长趋势，渗透率不断向上突破。展望 2025 年，考虑动力/储能/出口等不同链条，我们认为锂电需求仍有望保持较强增长动能。

汽车行业：全球汽车市场电车渗透率仍有较大提升空间－海外汽车市场分析报告-财通证券

汽车座椅行业：“第三空间_概念推动下，座椅迈向全维度创新阶段－深度报告-国元证券

汽车与零部件行业：人形机器人将打开汽零公司新的成长空间－人形机器人系列报告之一-东方证券

新能源汽车产业链行业：2024全球动力电池装车量TOP10出炉，当升科技中伟股份达成战略合作-东方证券

The global automotive manufacturing industry is dominated by large, multinational manufacturers including Toyota, Volkswagen, Hyundai, and General Motors. The leading players in this industry are tapping into the growing demand for hybrid & electric vehicles, while alliances within the industry are a common strategy. The automotive industry is continually facing new challenges, with globalization, individualization, digitalization, and increasing competition putting pressure on the industry. Moreover, companies are facing more stringent regulations in terms of their safety and environmental commitments. The competition in Electric Vehicles (EVs) and automated driving is growing more intense, yet it is conceivable that these technologies might be considered commonplace in 5 or 10 years. Beyond contending with their conventional rivals, these companies must also be adaptable to the emergence of startups and unconventional players in the automotive sector, including ride-sharing services.

• Autonomous vehicle makers are working hard to address the issue of autonomous vehicle safety when they are deployed on public roads. These players believe that autonomous vehicles will be far safer drivers than human drivers. This standard necessitates a large increase in existing testing installations.