Currently, the global automotive industry is undergoing profound changes, and with the help of rapidly developing artificial intelligence (AI), autonomous driving has become a key core technology "highland" for major car manufacturers worldwide. How far are we from fully autonomous driving now? What are the challenges to the practicality and security of this technology? These issues are receiving increasing attention and discussion from consumers. Experts say that autonomous driving is gradually realized along with the dynamic evolution of the vehicle's entire usage cycle. When a new model is launched and delivered, its autonomous driving function is not fully integrated. In general, the autonomous driving function relies on sensors such as cameras, millimeter wave radars, and LiDARs installed on the vehicle to obtain road information, which is integrated and fused into a vehicle centered "bird's-eye view" of the road conditions through an onboard computing platform. The vehicle's autonomous driving algorithm will use this to "infer" the corresponding driving path. After desensitization, the relevant data will also be uploaded to the cloud computing platform through the Internet, "feed" the AI big model for training, continue to iterate and upgrade the algorithm, evolve a new version, and then push it to the user's vehicle to continuously optimize the vehicle driving experience. According to the standards established by the International Association of Automotive Engineers, autonomous driving in a broad sense is divided into six levels from L0 to L5. L0 only provides warning information and does not intervene in driving operations. L1 and L2 are still primarily driven by the driver, and it is more accurate to refer to them as assisted driving. Only at L3 and above can it be considered as gradually decreasing until driver intervention is eliminated in autonomous driving. However, the higher the level, the more advanced the actual experience may not be. Wang Xin, Executive Vice President of Mercedes Benz (China), said that when the L3 autonomous driving starts, the driver's hands can detach from the steering wheel and focus on other things, but it must be within the corresponding operating design domain (ODD). "In practical use, L2 autonomous driving may have a wider range of applications than L3, such as on highways and urban and rural roads, but in these cases, the driver is still responsible." Pan Jiahui, Public Relations Manager of Bosch Intelligent Driving Control Business Unit, explained. "Let AI help me drive" is widely believed in the industry that at the current level of technology, it is still difficult for cars to get rid of drivers and become fully autonomous. Li Bin, the founder of NIO, stated that autonomous driving is still in the stage of "human vehicle co driving", and "from having this technology to truly being useful, it is going through such a stage.". "I feel like I need AI to help me drive, not just drive for me." Bai Guolong is a traditional fuel vehicle user who, after experiencing the autonomous driving function of a certain model, stated that the vehicle can indeed cope with the vast majority of common road conditions, but when there are road conditions or driving actions that do not meet his expectations, he will take over decisively. Ouyang Chen, the person in charge of the automotive industry media "Electric Planet", said that the current autonomous driving function still cannot match the efficiency of drivers when passing through complex road sections, such as encountering a large amount of pedestrian traffic at zebra crossings during turns. At present, autonomous driving technology still faces some technical bottlenecks and difficulties. For example, the Automatic Emergency Braking (AEB) system can theoretically assist vehicles in automatically stopping in emergency situations, avoiding collisions. But Pan Jiahui said, "This function has corresponding triggering conditions. When the driver rotates the steering wheel in distress, or the vehicle is driving on a larger curve, or the speed exceeds the system defined threshold, AEB may not be triggered.". In addition, identifying and avoiding sudden low-speed, stationary targets or irregular vehicles during driving is also one of the difficulties in the industry, and it is currently difficult to achieve 100% recognition and avoidance. Wang Xin stated that in the face of complex road conditions, especially the "corner cases" that arise during special road conditions, autonomous driving needs to accurately "infer" a safe driving path. "This also requires continuous improvement in algorithms, computing power, and effective data training.". It will take some time for fully autonomous driving to become a reality, but the auxiliary driving function has attracted the attention of many consumers. "Liberating energy and reducing accidents are all benefits that intelligent (assisted) driving brings to our users." Li Bin said that in terms of NIO's current products, the safety of co driving between people and vehicles has increased by 6.26 times compared to driving alone, and this aspect of performance is still improving. When activating assisted driving, the driver does not have to constantly press the accelerator pedal or be ready to brake at any time, which can free up a lot of energy. "The difference between co driving and self driving is that one is watching the car, and the other is staring at the car." Wang Xin explained that the various sensors equipped on the vehicle are equivalent to having many pairs of eyes to help the driver see the road, and can provide timely warnings in case of risks. During long-distance travel, co driving of people and vehicles can greatly alleviate driver fatigue. However, Li Bin also believes that consumers are "somewhat unfamiliar with assisted driving and do not know how to adapt to people and cars driving together. There needs to be a process of acceptance, and gradual progress is also reasonable.". Experts believe that overall, the standardized use of assisted driving functions has greatly improved people's travel experience, but no matter how advanced autonomous driving technology is, travel safety should always be given top priority. (Lai Xin She)