Technological Supremacy Chapter 272: 4 hardware problems!


After seeing off An Ran and Li Moran, Luo Jia called Shen Lang to her side after school in the afternoon, and asked him about his recent studies.

At the beginning, geniuses like An Ran and Cao Yuan all completed the six-year middle school courses in one or two years. The geniuses of Star University studied more intensively, and the teachers were more professional, so the progress was rapid.

By next spring, Shen Lang and the best genius in the school will finish their high school courses.

The slowest progress is those young foreign geniuses. They have to master Chinese skills while learning knowledge, so the progress is much behind.

However, Star University teaches students according to their aptitude, and each student’s schedule and schedule are different. These situations are completely under the school’s control and will not cause any trouble to the teaching work.

In general, Star University’s efforts to internationalize have failed in the end. Due to the resistance of foreign forces, the number of young foreign talents is extremely small.

Shen Lang frowned and thought for a few minutes, and then said to Luo Jia: “I think it’s very right for the teacher to fight back with great fanfare.”


“Because of faith.”


“Yes.” Shen Lang said seriously, “The entire domestic auto industry has made considerable sacrifices. The joint venture has not joined the Electric Vehicle Alliance, and due to the impact of the new energy plan, the entire China is Focusing on whether the auto industry can grow bigger, after all, only when the auto industry grows bigger can the problem of power layoffs be truly solved.”

“Simply put, electric workers are about to lose their jobs, and their nerves are already very fragile. We can’t let them lose confidence.”

Luo Jia nodded slightly. Although Shen Lang is young, he sees things very clearly. If others are laid off, they will be blamed. eat.

Therefore, on the road of expanding the automobile industry, Xingchen Technology must not give up even a single step, and must always maintain a strong position and give everyone full confidence.

That night, Luo Jia thought for a long time. When the brainstorming group gathered in the morning, he already had a comprehensive plan in his mind.

Many people who don’t understand autonomous driving technology will feel that this technology is nothing special, and they don’t know that in order to truly realize autonomous driving, Star Technology must solve four major problems at the hardware level alone.

The first problem is called industrial computer.

To put it simply, the industrial computer is the brain of the automatic driving system. You can also understand it as a small computer installed in the car body. No matter how advanced the algorithm is, no matter how much real-time data, it needs this brain to process it. .

At present, the most powerful Baidu Apollo system in China uses a GPU industrial computer, model Nuvo-5095GC.

Semiconductor Division of Star Technology took over the R&D task of industrial computer.

Fang Yuanchu suggested to Luo Jia that due to time constraints and heavy tasks, he could join hands with the team of his old company, which is HiSilicon Semiconductor under Huawei, to assign some development tasks to HiSilicon, and finally Xingchen Technology will do a full integration.

This proposal was accepted by Luo Jia. There are many semiconductor design companies in China, but only Hisilicon can rank among the top five in the world.

As for Star Technology’s semiconductor business unit, although it has been recruited for many years, its strength is considerable, but because the semiconductor business unit is not independent, they are still a mystery to the outside world.

Industrial computer has a lot of applications in the industrial field, but it also has huge levels. High-end industrial control and low-end industrial control have never been the same thing.

Just like the most advanced fourth-generation stealth fighter, one of the core technologies is the flight controller, and the flight controller is also an industrial computer in essence, but the technical level is really high. The national team has used a team of tens of thousands of people and has been working on flight control technology.

The industrial computer required to realize automatic driving is of course inferior to that of an airplane, but it is also a typical high-end technology. The entry of HiSilicon can speed up the development progress.

With the relationship between Xingchen Technology and Huawei, it is not difficult to attract HiSilicon. In terms of benefits, Luo Jia has never exploited the precedent of partners, and has always led everyone to make a fortune together.

As for the second problem, it is called CAN.

The interaction between the industrial computer and the car chassis must pass through the special CAN language, so CAN can be understood as an industrial language decoder.

For example, the current speed of the car is 60, and the sensor has detected the speed per hour, but how to tell this information to the industrial computer? Everyone is a machine, and no one can speak.

So CAN technology is needed to solve the problem of information pathways between machines.

The third problem is global positioning technology and inertial measurement unit.

Global positioning technology is not difficult to solve. Huaxia has the Beidou system. If it is combined with the North American GPS system, it can form a dual positioning channel to obtain accurate vehicle positioning information.

But the inertial measurement unit, or IMU for short, is more difficult.

Relying on global positioning and IMU, you can know where the vehicle is and which direction it is driving. In addition, IMU can also provide richer information such as yaw rate, angular acceleration, etc., which are helpful for the positioning of self-driving cars. and decision control.

The fourth piece of the puzzle is the sensor.

In current self-driving vehicles, three sensing systems need to be used at the same time, namely vision sensor, laser sensor, and radar sensor.

The visual sensor is a binocular camera, which can provide stereoscopic vision perception, not only for autonomous driving technology, but also for future robot technology.

Industrial binocular sensing cameras are not the same thing as mobile phone cameras, and the biggest player in this field is Star Optics, a subsidiary of Star Technology.

After years of expansion and development, Star Optics is no longer a small factory that can only manufacture cameras, but a large-scale comprehensive company with sensing, remote sensing, laser, astronomical optics, medical optics, optical software, etc. body.

There are mainly three foreign competitors, Mobileye of Israel, PointGrey of Canada, and Pike of Germany. Their technical level is lower than that of Star Optics.

Laser sensors are divided into single-line and multi-line. For each additional line, the cost will increase accordingly, and the detection effect will also increase accordingly.

At present, the highest level in the global civilian field can achieve 128 lines, and in the military field, it is no problem to achieve thousands of lines because of the cost.

Well-known laser sensor manufacturers include Velodyne and Quanergy in the United States, Ibeo in Germany, and Sagitar Technology, Hesai Technology, and Star Optics in China.

The last is the radar sensor, which is relatively common. The front and rear reversing radars are typical radar sensors, and the main researches on this thing are auto parts manufacturers, Denso, Bosch, Continental, Delphi and so on.

Star Optics did not make radar sensors, but another subsidiary, Star Electric, has a radar sensor team and production line.

It can be seen that although the sensor is the most complex and the most diverse of the four hardwares, it is the natural strength of Xingchen Technology. Xingchen Optical has vision and laser, Xingchen Electric has radar, and it can be used directly , no separate development is required.

So far, the tasks of the hardware department have been arranged. Fang Yuanchu’s semiconductor business department is leading, and Huawei HiSilicon and Xingchen Research Institute are cooperating.

Luo Jia glanced at the words written on the whiteboard, and said to everyone, “On November 1, 2017, the global Internet giant Google officially gave up the development of autonomous driving technology, and its subsidiary WAYMO was completely closed.”

“After nearly ten years and billions of dollars spent, Google still chose to give up. It can be seen that autonomous driving technology is actually very difficult, and the technical level is also very high.”

“In addition, there are Apple, IBM, Microsoft, etc. These giant companies have autonomous driving research teams, but who has launched a real product? No, not a single one.”

“Because autonomous driving not only needs to solve hardware what is more difficult to solve is software problems.”

“Fortunately, when we launched the electric balance car two years ago, we started the pre-research work on autonomous driving technology, and we also have the world’s largest industrial software team.”

“The software center that Li Moran is in charge of has now reached a scale of nearly 20,000 people. If it is not enough, we will also pull out the subsidiary Star Search! That is also a top team of more than 12,000 people!”

“In the field of software, we are also the strongest on this earth!”

After listening to Luo Jia’s words, everyone puffed up their chests proudly.

No one will forget that the software army of Xingchen Technology once won the three major epic projects of search engine, operating system, and industrial software. made history.

It can be seen from this that although the software army of Xingchen Technology is low-key, its strength is not low-key at all.

“After gathering all our forces, what we need to solve is the five major problems of the autonomous driving software.” Luo Jia paused and said.


Leave a Reply

Your email address will not be published. Required fields are marked *