电动车公社·2026-02-26 16:06Core Viewpoint - NIO has made significant advancements in its NWM 2.0 autonomous driving system, showcasing improvements in driving capabilities and user experience after a six-month hiatus in updates [1][3][59]. Group 1: Technological Advancements - The NWM 2.0 system has transitioned from a rule-based model to a world model combined with closed-loop reinforcement learning, allowing for more sophisticated scenario simulation and self-optimization [4][6]. - The system can now simulate driving scenarios and execute actions based on feedback, enhancing its ability to navigate complex urban environments [7][12]. Group 2: Performance Improvements - The narrow road navigation capability has significantly improved, allowing the vehicle to maneuver effectively in tight spaces while maintaining a speed of over 10 km/h [8][16][21]. - The system demonstrates enhanced decision-making skills, such as yielding to oncoming vehicles and utilizing available space for maneuvering, reflecting a deeper understanding of the driving environment [11][20]. Group 3: User Experience Enhancements - The NWM 2.0 includes a new feature for navigating to battery swap stations, which is more relevant for urban driving scenarios compared to previous versions focused on highway use [28][30]. - The system can autonomously handle tasks like identifying parking spaces and waiting in line for battery swaps, even in unfamiliar locations, enhancing user convenience [31][33]. Group 4: Areas for Improvement - Despite advancements, the system still exhibits some limitations, such as difficulty recognizing text-based waiting zones and occasional hesitation in decision-making, which may lead to user discomfort [44][46][54]. - There were instances of the system requiring driver intervention, indicating that while improvements have been made, further refinements are necessary for a fully autonomous experience [55][62].