Science and Technology Foresight ›› 2025, Vol. 4 ›› Issue (2): 144-157.DOI: 10.3981/j.issn.2097-0781.2025.02.011

• Review and Commentary • Previous Articles     Next Articles

Development and Outlook of Intelligent Driving Technology

LI Shengbo1,(), JIANG Kun1, TIAN Ye2, CHEN Chen3, SUN Jian2, YANG Diange1   

  1. 1. School of Vehicle and Mobility, Tsinghua University, Beijing 100084, China
    2. College of Transportation, Tongji University, Shanghai 200092, China
    3. College of Metropolitan Transportation, Beijing University of Technology, Beijing 100124, China
  • Received:2024-12-11 Revised:2025-03-27 Online:2025-06-20 Published:2025-06-26
  • Contact:

汽车智能驾驶技术发展与趋势展望

李升波1,(), 江昆1, 田野2, 陈晨3, 孙剑2, 杨殿阁1   

  1. 1.清华大学车辆与运载学院,北京 100084
    2.同济大学交通学院,上海 200092
    3.北京工业大学城市交通学院,北京 100124
  • 通讯作者:
  • 作者简介:李升波,教授,博士研究生导师。获国家高层次领军人才、教育部青年科学奖、交通运输行业中青年科技创新领军人才等。人工智能国际评测组织MLPerf自动驾驶咨询委员会委员、IEEE智能交通系统学会的Board of Governor委员、中国汽车工程学会青年工作委员会主任委员(首任)、中国汽车工程学会人工智能分会主任委员(首任)等。主要从事强化学习、最优控制、神经网络及其在自动驾驶和机器人领域的应用研究。获国家科技进步奖二等奖、国家技术发明二等奖、中国汽车工业科技进步特等奖、中国自动化学会自然科学一等奖等。发表论文200余篇,H因子78,入选IEEE期刊封面论文5篇,国内外学术会议优秀论文奖12次,连续4年入选爱思唯尔中国高被引学者。电子信箱:lishbo@tsinghua.edu.cn

Abstract:

Intelligent driving technology has the potential to enhance the safety and efficiency of road transportation systems, drive innovation and technological breakthroughs within the automotive industry and its upstream and downstream supply chain ecosystems, and serve as a critical area of competition among leading industrial nations. After decades of development and with continuous advancements driven by the Internet and artificial intelligence technologies, intelligent driving technology is approaching the threshold of large-scale commercial application. This paper provides a systematic review of the development trajectory of intelligent driving technology, summarizes the current state of technological progress in hardware support systems, software functionality systems, and “end-to-end” autonomous driving systems, analyzes key technical issues and challenges faced by various functional modules, and offers insights into future development trends. The findings aim to provide scientific guidance and strategic recommendations for the further development and application of advanced autonomous driving technologies in China.

Key words: intelligent driving, “end-to-end”, perception and prediction, decision-making and planning, simulation and testing, closed loop of data

摘要:

智能驾驶技术作为提升交通安全与效率、推动汽车产业革新的核心领域,已进入高级别自动驾驶量产应用的关键阶段。文章系统梳理了智能驾驶技术的发展历程,从硬件支持系统(车载芯片、传感器、控制器、执行器等)、软件功能系统(操作系统、地图与定位、环境感知、预测决策、运动控制、仿真测试等),以及“端到端”自动驾驶系统(数据采集和标注、训练算法、模型架构、压缩部署等)3方面总结技术现状。揭示其面临的共性挑战:硬件能效与安全瓶颈、传感器融合精度不足、开放道路长尾场景泛化能力弱、大规模模型算力需求激增、仿真测试覆盖不足及数据安全风险等。未来技术将向硬件融合一体化、软件高稳定性与安全性、“端到端”车路云协同赋能方向演进。针对中国发展需求,提出加强硬件自主可控、构建国产软件生态、建立统一测试评价体系及推动“端到端”技术落地的对策建议,以期为高级别自动驾驶技术突破与产业化提供理论支撑。

关键词: 智能驾驶, “端到端”, 感知预测, 决策规划, 仿真测试, 数据闭环