At this stage, driving assistance still belongs to L2 level. For me, my concern is:

1. There are more and more lidar and more hardware. Is redundancy necessary?

2. Some autopilot functions are still the future. Is it wise to pay for the hardware cost at this stage?

3. Is it really good to experience the hype function of the enterprise?

Due to the cost and the maturity of the algorithm, at present, no automobile manufacturer can realize the so-called automatic driving, and the name of ADAS system only uses "assisted driving" once. The so-called "assistance" means that the decision-making power and initiative of driving should be completely in the hands of the driver. Therefore, the ADAS functions installed on all production models at present, including lane keeping, cruise control and even automatic lane change, are not designed to completely replace the driver's operation, but to reduce the driver's operating pressure.

Some car companies began to differentiate competition, high-speed-parking lot-urban area, and began to draw circles to expand driving assistance application scenarios. If the experience is good, this is the function that consumers can actually see and touch. To tell the truth, Li Yanhong's speeches in recent years are relatively conservative. At least, he won't brag in public, basically according to the conservative statement. Therefore, Baidu's autonomous driving is not ahead of Tesla, which is a technical problem. It takes too much professional data for comparison to draw a conclusion, which is hard to say.

Let's put facts and reason. Before Baidu's self-driving, most of them were in Apollo taxis, but they were self-driving cars where no one really sat in the driver's seat and then the steering wheel turned by itself. At present, there are almost100000 people, and it is really necessary to take a taxi. This is not a test drive. As an ordinary person, if you go to the next radish to run the App, you can get a taxi. You can try it. The experience is quite online.

Tesla's autonomous driving still requires someone to sit behind the steering wheel. Although it can also park automatically or drive remotely, it still needs the driver's participation if driving on the city streets. There was no accident before, but Tesla appeared in the advertisement with the words "The system is designed for short-distance and long-distance driving without any action by the driver". As a result, it was sued by the California Motor Vehicle Authority with misleading labels and statements. Later, Tesla also issued a disclaimer, mainly warning that these functions still need the active supervision of the driver. Therefore, Baidu and Tesla, in terms of autonomous driving, can almost get a glimpse.

To be honest, Du Ji is likely to be ahead of Tesla, but it will take time. Start with Tesla. Their false propaganda about autonomous driving is true. Tesla's advertisement says how convenient it is to drive even if your hands are off the steering wheel. In fact, Tesla does not support leaving the steering wheel with both hands. Their so-called "automatic driving assistance system" is used to realize automatic assisted navigation and driving on the expressway. And assisted navigation and autonomous driving, here are drivers.

What is autopilot in Du Ji? Let's just say that Baidu has just obtained the first batch of unmanned demonstration operation qualifications from Chongqing and Wuhan municipal governments. Du Ji's auto-driving is not only an auxiliary navigation, but a real unmanned auto-driving. Therefore, as Li Yanhong said, from the perspective of autonomous driving, Tesla really has to be caught up by concentration. Even at the level of autonomous driving, Tesla has not yet reached the level of concentrated cars.

But regardless of the level of autonomous driving, Tesla is still very popular. First of all, the brand has great influence. As an imported car, the brand Tesla is very popular in recent years. Followed by the model, SUV is the most popular model in China. Take Tesla Model Y as an example. As a medium-sized SUV, it naturally meets the market demand. Because of its longer time to market and more experience, Tesla will have a deeper understanding in some aspects. Therefore, on the whole, Du Ji Auto Intelligence is a potential stock, but it will take time to surpass Tesla.

Finally, Li Yanhong said, "The midfield of smart cars is electrified and the end is intelligent." I think intelligence is only the end of smart cars. Just as we didn't believe that cars could be driven by people before, now we have realized automatic driverless driving. . Then, can we boldly imagine that the future car will fly and the future car will be powered by air? In any case, there is no end to development, and I believe that smart cars will have the next stage.

The "dual-system" autopilot solution of "pure vision+multi-sensor fusion lidar" has been adopted, and NVIDIA's dual-chip 508TOP supercomputer system and 8295-chip intelligent cabin system have also been put on board. The dual radar design can realize the horizontal FOV coverage of 180 degrees, identify the key areas with dense vehicles more accurately, and greatly improve the safety of automatic driving in typical long-tail scenes such as queue jumping and pedestrians crossing the road.

So I think there is no problem with Du Ji's autonomous driving technology, but there is still a problem that it is delivered to the road, and its application in specific practical scenes is still relatively small. Compared with Baidu's driverless taxi drivers who run a fixed road every day to collect data and correct algorithms, unmanned vehicles will undoubtedly face more complicated road environment and corresponding road regulations. In short, I think there is still a long way to go.

From Higaja to Metauniverse, from sci-fi words to robot female voice of "Hello, I'm a car robot in Du Ji", Du Ji is trying hard to establish a sense of scientific and technological experience. Whether it is an early flash car with intelligent elements, or the first demonstration of assisted driving ability on urban roads, or even the sharing of the concept of intelligent mobile at the "car robot" and Baidu AI conference, it is firmly buckled with the enthusiasm of science and technology, giving people a strong scientific and technological color rendering.

Different from other OEMs and brands' introduction to driving, the car itself and services, Jiji is almost very focused on the research of the technology itself (at least for now), and continues to provide content exposure in the field of science and technology, and directly announces that it can provide complete intelligent driving and intelligent cockpit capabilities when delivering the car. This is almost the first brand to make such a commitment in recent years. I think this is also a response to whether the market can land Baidu's automatic driving ability.