Rewriting the Technological Scenario

After Mask left, Meng Qiao directly about BYD's Wangchuanfu and Geely Li Shufu, because many things were not waiting, the two companies were very frequent. Baidu search, more look at free reading.

"I heard that BYD and Geely began to increase the investment of electric vehicles."

"This year, Omei vigorously promoted electric cars. After the Copenhagen conference, electric cars became the topic of the industry's hot discussion, and Meng always known, I have always believed that the electric car will be tomorrow." Probably because Li Shufu is present, Wang Chuanfu said obvious keep it.

In fact, the two knew that it was a little accident while being approved. This will gradually wait. It is a opponent. Even if the wind group is a big shareholder, Meng Qian said that today is to talk about future planning, competitors Sitting together to talk about future planning, how do you feel unsatisfactory.

Sure enough, Li Shufu is only a very official guest package. "In the development of electric vehicles, we still have many places to learn to BYD."

I know that the two are careful, so Meng Qian direct theme, "I found two times today, I first wanted to tell two I am preparing to reconcile with Mask."

Meng Qian probably told things.

"Tesla is really valuable to R & D in the motor system this year." About Tesla's technology, Li Shufu is still more comfortable, after all, Tesla is in the pure electric car, Geely did not develop this direction. .

But Wang Chuanfu is clearly a bit unsure. "In fact, we have many patented overlap with Tesla, and recently accelerated the practice of patent applications.

"It is actually because of the development of BYD, I think of Tesla's patents, so two sides combined, we have the opportunity to impact more leading technology."

The two nodded, recognized Meng Qian's statement, and then waited for Meng Qian, after all, they didn't know what Meng Qian was looking for him. As a result, Meng Qian directly transferred the topic. "This thing is to follow The two first greeted, and the specific situation, I will say that I will talk to Mask to reach the final agreement, and the specific development of electric vehicles will eventually rely on two. Baidu search, more look free to read.

I am looking for two in order to find another thing today. This year we have seen an electric vehicle a trend in the development of automobiles, and there should be another trend of car development. "

Wang Chuanfu and Li Shufu have opposed his eyes and did not think of the answer, and the last topic was so embarrassed, in the face of the eyes of the two, or Meng Qian gave an answer, "drone."

When I heard the answer, the two suddenly realized.

"This year, Google has already officially laid out no one, and it is backed by the National Defense Advanced Research Project Bureau. The large number of R & D staff in the bureau went to Google. In this field, I didn't want to go behind." Meng Qian simply fell.

"What is the specific intement for Meng Tong?"

Meng Qian stigted two people to see PPT display, "Our company has taken out a conclusion, sensor + high precision map + cloud computing + rational decision algorithm is the most likely to achieve unmanned scheme.

Therefore, if you want to develop unhaly, you need a corresponding lens, corresponding C touch S image sensor, corresponding DSP digital processing chip, corresponding millimeter wave radar, corresponding laser radar, etc. Hardware equipment, and a rational decision-making algorithm and A safe high-speed cloud platform.

Therefore, for traditional car companies, no one is actually a new field, although the most important thing is still safe, but traditional car companies are all functional security routes, in the unmanned area, functional security is far from enough.

But is no one driving is a new industry? I don't think that when no one is really mature, he will inevitably hit the traditional car company. If the user trusts no one drive, it is willing to drive people will reduce.

So I want to ask two times today, is there interest to be involved in this car enterprise revolution with us. "

Both people understand why Meng Qian called them, Meng Qian wants to engage in no one, do not want to put the bet on a traditional car company, so I call the two together completely no problem.

"When Meng is preparing to enter a industry, it must have been prepared?" Wang Chuanfu said very certain to Meng Qian's understanding.

Meng Qian smiled, indicating to continue to see PPT, "Just said, no driving requires laser radar and millimeter wave radar, because laser radar can accurately measure the relative distance between object profile edges and devices in the field of view, accuracy It can reach the centimeter level, and for the shortcomings of laser radar is not enough, millimeter wave radar can make up.

Therefore, we have made a lot of research around the laser radar and millimeters, for unmanned driving, we have developed this 64-wire laser radar, two views of this dynamic map.

The laser beam is emitted from both side lenses. After encountering an obstacle fold, pass through the intermediate lens, the receiver can determine the obstacle by processing analysis.

All time information, control information, both the received signal transmitted to the top of the top, for data analysis, and the bottom rotation recorder can record the position and rotation information when rotating, all raw data transmitted to the computer through the bottom data line.

As for the greatest advantage of this product, 64 beam radial injected in the vertical direction can cover the angle of the vertical direction 26 °, the detection distance of 100-120 meters, the detection barrier is around about 10m, and the laser radar is The second rotation is 10 times, and the horizontal 360 ° horizontal coverage can be achieved, and the output points are as high as 130,000 information. "

This thing is still able to understand and know more and more digestive, but Meng Qian will make two people feel a little bit, "But in my opinion, our biggest advantage is in algorithm .

As far as I know, several unmanned enterprises, including Google, driving training using deep learning, directly exporting the output of the controller from the input of the sensor, however, as the founder of the depth learning, I can basically confirm this depth Learning is a wrong direction in the unmanned area because it belongs to correlation reasoning, not due to fruiting.

For example, people will slow down in the rain, this deep learning logic is going to slow down, but it doesn't know why it is going slowly, then we don't give it to snow, there will be a situation. Unmanned driving will slow down in the rain, but driving in the normal speed of the sky, and may even be driven by other data.

This algorithm for this unknown cause is a great hidden danger for driving safety, because there is too much unknown situation during driving, and it is necessary to make a causal judgment.

Therefore, we need to introduce new decision-making mechanisms in the automatic driving area, we found such a mechanism, which is the Bayesian network, the Bayesian network is a probability reasoning system, the Probability of the Bayesian network for events in data processing As well as event credibility analysis has a good classification effect, it has two decisive advantages: modular and transparency.

The modular advantage is especially important, such as updating the car gearbox, when the gearbox is replaced, we don't have to rewrite the entire drive system, only need to modify the subsystem of the transmission model, the rest can remain unchanged. .

The company with a large number of driving data analysis using the Bayesian network, as far as I know, there is only one of our wind groups.

And this is our great advantage. "

I don't know why, Meng Qian took two leading technology in Wang Chuanfu and Li Shufu seems special normal, and even there is no more than half.

Meng Qiao is actually a clear attitude of the two. Seeing that they have no resistance to the target. In 2010, the Great Wind Group should carry out a driverless experiment of at least 100,000 kilometers, actively grabbing people's driving technology highland.