Director Li was shaking his head and sighing, but Chen Yao was very leisurely. After dinner, he sat on the sofa in the living room and drank tea.
Qin Yunzi got up and went to her room to take a shower and rest. Unlike a night owl like Chen Yao, she went to bed very early every day.
And Lin Yuhan is sitting in the living room with her notebook, coding and writing a novel. The last book she wrote has been sold to Li Chenguang. There is a high probability that the book she is writing now will be turned into a film and television series.
Suddenly, Chen Yao¡¯s cell phone rang. He put it to his ear and heard the voice of Professor Zhong Guoqiang. He and his team were testing unmanned vehicles nearby.
While doing a test on a city road, Chen Yao asked for his address, then stood up and walked out of the house wearing flip-flops.
Park nearby.
Wearing slippers and lazily arriving at the park, Chen Yao saw Professor Zhong Guoqiang and his team at first sight. A Zhanlong electric vehicle was parked on the roadside.
Even though it¡¯s very late now, Professor Zhong is still working. It¡¯s really hard for someone in his sixties.
Chen Yao poured a bottle of ice fountain for him: "Professor, thank you for your hard work."
Zhong Guoqiang smiled lightly: "I have always been a car fan, and it is an honor for me to personally participate in the development of the Zhanlong unmanned vehicle."
Zhanlong Automobile is conducting night tests to test the safety of driving at night under unmanned autonomous driving.
There was no one in the sky-blue car body. The Zhanlong car had just circled the east, south, west and north of the park and returned to the far point.
Chen Yao said: "What level has Zhanlong Automobile reached now? Is it still L3?"
Zhong Guoqiang responded: "It's close to l4."
Autonomous driving is divided into five levels, from L1 to L5.
L1 is just the simplest assisted driving, relying on adaptive cruise.
You can only release your feet, but your hands cannot leave the steering wheel, and the driver still has to control the car.
L2, semi-autonomous driving, but the driver still has to keep an eye on the driving situation of the car.
This kind of semi-autonomous driving is the most common so-called driverless car that has been mass-produced.
When it comes to L3, it¡¯s highly automated driving.
Drivers don¡¯t have to keep an eye on the driving situation. They can play games on their mobile phones and do intimate things with their lovers. However, in certain situations, people still need to control the car.
In other words, a car of this level cannot fully rely on its autonomous driving, otherwise it will be a joke with one's own life.
Some companies have already achieved mass production of L3-level cars and launched them on the market.
When it reaches the L4 level, companies all over the world are testing it, but no company can do it yet, and no model has been launched on the market.
This is a higher level of autonomous driving. The driver does not need to monitor the operation of the car. Not only does he release his feet and hands, he no longer even needs to look at the road surface.
Therefore, if you want to achieve driverless driving, you must at least achieve L4 level.
Now, the Zhanlong built by Professor Zhong Guoqiang and his team is an L4 car. Of course, it is still just a test, and there are still many problems.
If L5 is really achieved, it will be fully autonomous driving. Currently, L5 cars are just a legend.
No company in the world can do it, let alone mass production, there is not even a test car, it is just a beautiful fantasy, something at the level of black technology.
Although L4 can also achieve driverless driving, it can only operate on specific road sections, specific areas, and specific cities.
And the L5 car can handle no matter how complicated the situation is, and it is safer than driving a car.
If you want to realize L5 autonomous vehicles, the key lies in sensors and algorithms.
Zhong Guoqiang took a sip of ice water: "The biggest problem right now is the sensor. Do you think our new generation of cars should be equipped with lidar?"
Many L3 autonomous vehicles on the market are not equipped with lidar, and some have it installed but eventually dismantled it.
Even Tesla¡¯s founder, Boss Ma, said that radar is useless and not as reliable as a camera.
Zhong Guoqiang¡¯s R&D team encountered the same problem. Do we need radar? Not just laser sensors, but can millimeter waves, ultrasonics and the like be removed?
Chen Yao smiled and asked: "What do you think, are laser sensors useful?"
Zhong Guoqiang affirmed: "It must be useful, it's just a matter of how big or small the effect is."
The laser sensor can emit something called a "laser beam", which can be used to sense surrounding obstacles, distance, movement speed, etc., with accuracy down to millimeters.
Not only??, something that seems awesome has very obvious shortcomings.
It¡¯s too rigid and not smart at all.
A driverless car was driving on the road, and suddenly a big plastic bag flew over
Sorry, the lidar cannot identify it at all. What it is, I only know that it is an obstacle and how far it is.
This will lead to a very terrible consequence
Suppose that an unmanned vehicle is driving on the highway and a large plastic bag flies over. The laser sensor recognizes that it is an obstacle and issues an emergency braking command.
A large truck suddenly rushed up from behind and hit the rear of the unmanned vehicle.
It was a plastic bag, but it caused a serious accident.
Not only lasers, but other millimeter waves, ultrasonic waves and the like can only detect but cannot identify.
However, the camera can identify what the object is. It is just a plastic bag, so you can just drive past it without worrying about it.
Another reason why many unmanned vehicles are not equipped with laser sensors is cost. A laser sensor costs US$10,000.
It is not worth it to spend a lot of money for a sensor and increase the cost of the car.
In contrast, the camera costs only $5 and can do things that laser sensors cannot.
After Chen Yao thought about it carefully, he said: "Although the laser sensor is not perfect, it is useful after all and needs to be installed. But the camera recognition will be given priority."
In other words, the camera is the main force of driverless driving, and lidar and millimeter-wave radar are only auxiliary.
The camera is not perfect. Behind it is a set of AI algorithms. Since it is an algorithm, there will be errors. Especially the current artificial visual recognition cannot be fully as smart as humans.
Let the price of laser sensors come down, it still has great value.
Zhong Guoqiang asked: "How do you solve the cost problem?"
"I've decided." Chen Yao snapped his fingers: "We independently develop laser sensors, which not only have stronger performance, but also lower costs."
Since the debugging of the sensors on the market is not ideal, we might as well develop our own research and development of more advanced radars.
Chen Yao¡¯s lips curled up: ¡°I will upload the scientific research materials to the ¡®Super Quick Learning Platform¡¯ later, and let the scientific researchers go up and study, and they will be done soon.¡± (Remember the website address: www.hlnovel.com