Recently, the remarks of Mr. Tan Xuguang, chairman and CEO of Weichai Power, have caused great controversy. He said, "Driverless driving is a flicker, and assisted driving is hope." Driverless will be implemented first on mine trucks. Is this statement correct? Let me give you my own opinion.
First of all, I think his remarks are at least one-sided. I have pros and cons. First of all, what I agree with is that if we stand in the moment of 2022 and look at some companies advertising that their models are driverless, it is indeed a fool. Therefore, when many companies promote their cars, they will not use the 4 words of driverless driving, but use "advanced assisted driving" to replace. Therefore, whenever any company uses the words driverless when promoting its own car, it is absolutely false propaganda.
Why is unmanned driving a fool at the moment? Because the so-called smart cars of the moment just make the car smarter. Putting a chip on the car is equivalent to a human brain. Installing a camera on a car is equivalent to a human eye. Equipping the car with radar is equivalent to a human ear. And equipping the car with an advanced computing system is equivalent to giving the car itself the ability to make cognitive judgments. Thus achieving unmanned driving.
But what are the problems with driverless driving now?
The first problem is the need to learn, that is, the constant need for road tests to accumulate experience. It's like we humans. Teachers in class taught us to be polite to pedestrians, not to speed, and to obey traffic rules. These basic traffic norms are also mandatory in autonomous vehicle systems. But in real life, the road conditions we encounter are very complicated, and there may be animals or pedestrians who suddenly cross the road. Maybe when you have a red light in front of you, there is a 120 vehicle or a 119 vehicle in the back, so whether you go or not. When we encounter these problems, even our human driver has to think in his head, let alone the car, he does not know how to do it. Therefore, he needs to constantly conduct road tests, learn in practical road application scenarios, and accumulate experience. So many cities now give self-driving cars licenses that allow learning on specific roads.
Is this learning process difficult to learn? It is like this, because the traffic conditions are not the same in each different city, and in a city like Chongqing, the traffic situation is particularly complicated. And many third- and fourth-tier cities or rural areas. There are no clear lane lines. Then how the vehicle itself should judge also needs further learning.
The second problem is the problem of delay. We humans drive cars because we know in our heads how to walk the road, so what about self-driving cars, which are the navigation we set up in advance. Then self-driving cars, it is necessary to navigate in real time to guide the vehicle to drive. Navigation needs to know if the road ahead is congested, and even navigation needs to know whether you are in lane 1 or lane 2 in the lane, accurate to the lane line. And will the navigator also know if there is a red or green light 500 meters in front of you? So self-driving cars need precise navigation, even close to military-level navigation.
And now, the 5G network we use is not very popular, then in the case of 4G network. Navigation to guide you to traffic information is delayed. So in the absence of accurate navigation, our car machine system itself to give you navigation, he only knows where the destination is, will only foolishly follow the navigation to walk, he does not know whether the front is a red light or a green light. Therefore, it cannot predict the road situation in advance, which is more testing the vehicle's ability to judge the road conditions autonomously. And if we are entering a tunnel or a place with a bad signal. Then for self-driving cars, there are more problems.
In addition, we also have to take into account the weather conditions such as rain, snow and fog, even in the cold northern winter, if your vehicle is frozen and just blocks the camera or radar, then it can still achieve normal automatic driving, which is difficult to say.
So in 2022, in the present, saying that driverless driving is flickering, I am in favor of this view. But if we take a long-term view and say that unmanned driving is a fool, I don't approve of it.