Core Insights - The company has developed the industry's first high-performance large-scale reinforcement learning planner for autonomous driving, enhancing the capabilities of its unmanned vehicles in complex scenarios [1][2] - The research has been accepted at the prestigious CVPR 2025 conference, showcasing the collaboration between Zhejiang University and the company's autonomous vehicle technology team [1] - The new planner, named CarPlanner, addresses inefficiencies and performance issues in trajectory planning using reinforcement learning, outperforming imitation learning and rule-based methods on the nuPlan dataset [1] Company Developments - The company is advancing its logistics technology commercialization and international expansion, with L4 level unmanned vehicles now available for sale nationwide and operating in over 30 districts [2] - The autonomous vehicle technology team has collaborated with Nanyang Technological University in Singapore to develop a new end-to-end autonomous driving system (PAD), which has achieved the top ranking on the NAVSIM platform for autonomous driving simulation and evaluation [2] - The company has implemented automation and digitalization projects in over 600 locations across 28 countries globally [2]
持续布局“AI+物流” 菜鸟研发业内首个自动驾驶规划器
Zheng Quan Shi Bao Wang·2025-03-05 03:07