Expensive self-driving cars released by Tesla, Huawei, etc. no longer need high-precision map support (because roads are changing in real time, such as maintenance, so it is almost impossible for high-precision maps to accurately reflect actual road conditions in real time in theory, although they are predictable over long distances). May I ask: Is it difficult for the company to recover the high-precision map research and development that it invested heavily in the early stage? What countermeasures are you going to take in the future in terms of high-precision maps? From now on, is it a strat

NavInfo: Hello, the so-called non-dependence usually distinguishes functions between mapped and unmapped areas. Road sections without high-precision maps usually lead to functional degradation and manual takeover. Secondly, a method similar to crowd-source mapping will be adopted to actually use high-precision maps to support the realization of NOA functions, which does not mean that high-precision maps are really not used. With the development of autonomous driving, the real-time perception results are improving the cognition of the real world, but there are still a large number of regular or irregular long-tail scenarios, which leads to a higher dependence of vehicle control on maps, including scene judgment, positioning, rule-making, computing power saving, etc. The long-term existence of these challenges means that maps are necessary for a long time. Thank you.