Rebirth 08: Rise from copycat phones

Chapter 365 Robot Prototype
Xu Shenxue was really interested in the robots made by the robotics team, so as soon as the meeting ended, he asked them to take him directly to the laboratory to see the robots.

"Mr. Xu, the row of robotic arms in front of you are the various types of industrial robots we have developed and launched on the market. There are 23 types in total, each with different uses for intelligent production, which can replace most manual processes. Some of these industrial robots can also be used for precision operations. We have even tried to install these robotic arms in the laboratory to replace test personnel in various basic routine test operations, and they are quite useful!"

"What is being tested in the lab next door is our new generation of catering service robots, which are great at making coffee and milk tea."

Xu Shenxue took a look and found that all of them were typical industrial robotic arms. However, there were also several robots that were mounted on wheeled mobile platforms, but their core was still a robotic arm.

The flexible operation of the robotic arm is the core difficulty of current robotics technology...so most robots for work purposes actually focus on the development of these two robotic arms.

The same is true of the domestic robot that Xu Shenxue saw next.

Soon, Xu Shenxue saw the key project of the project team: the household robot.

There are two more types of household robots in front. One is an anthropomorphic bipedal, two-arm robot, which looks quite sci-fi and is in line with people's perception of robots.

The one next to it doesn't look as sci-fi, but it is actually more practical and is also the product that will be promoted. It is a wheeled robot. It is a body mounted on a wheeled mobile vehicle, and has two slender robotic arms.

"Mr. Xu, look, the one on the left is a bipedal, bipedal humanoid robot that we developed. Its arms are actually the same as the wheeled robot in terms of hardware and software. The difference is that it is equipped with a bipedal mobile platform in the lower part. However, we have not yet accumulated enough research and development experience in bipedal mobile platforms, so its mobility is a bit clumsy and not flexible enough!"

"Although it can meet basic mobility needs, such as running, jumping, climbing stairs, and descending stairs, it is still not enough."

"To be honest, this bipedal domestic robot is not mature yet and has not reached the stage of being put on the market. This release is just a conceptual release. We plan to release it commercially next year."

"We still need at least one year to accumulate more technology on the bipedal mobile platform, especially the accumulation of AI models, to solve and improve the many technical problems of the bipedal platform."

“What we are really going to release this winter is a household robot with a wheeled platform!”

"After replacing the bipedal platform with a wheeled mobile platform, the mobility problem can be solved instantly. Although the wheeled mobile platform lacks the ability to go up and down stairs and pass through complex terrain, the intended use scenarios of our domestic robots are at home, hotels and other places, and the wheeled mobile platform is already sufficient for initial use!"

"Our previous industrial robots and catering service robots all have wheeled mobile platform models. They are very easy to use and customers generally give good feedback."

"Mr. Xu, look, this wheeled housekeeping robot in front of you is the first generation of housekeeping robots we developed. It uses an electric wheeled mobile platform, 85 centimeters long and 65 centimeters wide. It has four wheels controlled by independent motors, and any wheel can move forward, backward, and turn. It also has an independent active suspension system that can provide high-precision dynamic balance for the entire platform."

"In theory, when the mobile platform maintains a speed of five kilometers per hour and passes through some slightly uneven hard roads, it can still maintain the levelness of the entire platform. If you put a glass of water on it, it can ensure that the water will not spill!"

"In addition, when passing through some necessary complex terrains, our robot can also be assisted by a pair of mechanical arms to achieve better passability!"

While the project leader was introducing the project, technical staff also started to operate and demonstrate!

The person in charge used voice commands to wake up the robot, and then gave instructions for the robot to perform a moving show, and even passed through complex terrain and climbed stairs! "

This feeling seemed quite strange to Xu Shenxue... When passing through some complex terrain on the ground, the robot's two robotic arms can extend out, touch the ground and provide a certain amount of assistance.

It's like watching an excavator use its bucket to perform complex ground operations!
That’s not all. Xu Shenxue also saw the housekeeping robot perform a very special sexual demonstration.

When it came to a suspended ground, it actually stretched out its mechanical arms and grabbed an iron rod above, hanging itself up... The two mechanical arms grabbed the levers alternately and moved itself over like a monkey.

The passability of this thing... is pretty awesome!

This seems easy enough, and humans and even monkeys can do it easily, but the problem is... the one opposite is a robot, and a wheeled robot at that. However, because it has a pair of mechanical arms, it has evolved various very special operations.

Who would have thought... that a robot that was supposed to walk stupidly on the ground could actually hoist itself up and swing around... This is even more awesome than the wheeled robots shown in many so-called science fiction films!
The wheeled robots in some science fiction films, the imagination of those directors, screenwriters, and prop masters is still a little lacking. Making a wheeled robot can only walk on the ground, and it will freeze when encountering stairs!
But this wheeled robot in front of us, with a pair of mechanical arms, has strong maneuverability...

The head of the robotics team continued:
"The wheeled mobile platform of this robot is transplanted from the wheeled platform of the wheeled industrial robot. From the hardware point of view, it is a very mature wheeled mobile platform!"

"On this wheeled mobile platform, we installed the core system of the domestic robot and two main operating robotic arms."

"The robotic arm is driven by a motor and has multiple joints that can move at a large angle. It can perform various operations very flexibly. The hardware technology of this robotic arm is also very mature. We have accumulated a lot of technology in the industrial robot and catering robot series."

"Of course, none of the above hardware is the core technology of our robot project. As for the hardware of these mobile platforms and robotic arms, many manufacturers can actually design and make them, but the hardware made in this way is just an isolated body."

"In the entire field of robotics, the most critical aspects are information perception, algorithm framework, and computing power core!"

"In terms of information perception, we have installed a miniature laser scanning system for this housekeeping robot, right here on the top of the robot!"

"This miniature laser radar is the latest scientific research product of our group. It can emit 128 lasers at the same time to scan the surrounding environment and create a 3D image, providing our robots with ultra-high-precision three-dimensional space perception."

"The robot is also equipped with multiple high-definition cameras, infrared cameras, ultrasonic radars and many other sensors. A very complex tactile sensor array is also installed in the fingers of the robotic arm."

"This tactile sensor array is a tactile sensor developed by us. It has more than 2,000 tiny sensors built in. It can accurately sense more than ten kinds of multi-dimensional tactile perceptions such as pressure, friction, soft and hard texture, and temperature. All of these combined can enable the hand part of our robotic arm to achieve ultra-high-precision multi-dimensional touch."

"The robot's hand is also the most complex part of the entire hardware platform."

"With the information support of this tactile sensor array, and with the support of our special hand activity algorithm framework, after special data feeding training, our robot's hand can already imitate most human hand movements, and can gently grasp some soft objects without destroying them."

While he was talking, the technicians in the front were also testing the robotic arm. The robotic arm picked up a banana, peeled it, and then held the peeled banana... The key point was that the banana was not damaged!
This control of force may not be a big deal for living things, but it is actually very difficult for machines.

"In order to support the robot's various actions, to collect external information, and to perceive the situation inside the machine, in addition to this special touch sensor array on the hand, we have deployed more than 300 different types of sensors throughout the robot to collect various data!"

"A large amount of sensor data is collected and then transmitted to the 'intelligent control center', which is the core part of the robot."

"For this intelligent control center, we use the latest EYQ3 computing chip, with a single chip computing power of 80TOPS. The entire control center uses four EYQ3 chips, providing a comprehensive computing power of 320TOPS for the entire system."

Although the EYQ series of computing chips were originally developed for autonomous driving of cars, they are not actually specialized autonomous driving chips.

It is essentially a computing chip used in the field of artificial intelligence, and is designed to process large amounts of sensor data at high speed... As for what the data is used for, it has nothing to do with the chip itself, it is a problem of the algorithm.

Therefore, the use of this chip is actually very wide. It can be used not only in smart electric vehicles, but also in drones and other unmanned equipment.

This is why a lot of secret unmanned equipment projects are basically tested with the EYQ series computing chips...it is really easy to use.

"The combined computing power provided by the entire system has surpassed that of the latest Hailan car, and we have also equipped it with a large number of low-power auxiliary computing chips for computing power processing of special parts and operations."

"For example, for data processing of the hand, because the amount of data is too large and the data is complex, we specifically put forward the requirements and asked the semiconductor department to design an AI processing chip 'ZY23' based on the -nanometer process for us. This chip is specially designed for the touch array of the hand and can process a large amount of tactile sensor data very efficiently and provide feedback."

"The comprehensive computing power of the control system, coupled with the auxiliary computing power provided by a large number of auxiliary special chips, enables our entire robot system to have a larger data processing capacity, and good data processing capabilities can also make the system more flexible and smarter!"

“When we combine complex and sophisticated hardware, self-developed high-performance sensors of all kinds, with the top computing platform, and the robot model that we painstakingly developed and trained with great effort!”

"Finally, we got this housekeeping robot!"

"Although it is just a robot for housekeeping purposes, it is actually the most complex and sophisticated high-tech equipment our entire group has ever manufactured!"

"Every component and every line of code inside represents the most advanced and cutting-edge technology of today's mankind!"

As the tester spoke, the robot in front began to do housework in a simulated room under the tester's instructions, mainly cleaning the simulated dining table.

Xu Shenxue watched the robot flexibly use a pair of mechanical arms to put away the dishes, pour the remaining kitchen waste into the trash can, and then use the dishwasher to wash the dishes...

Although the robot's movements were relatively slow, it still made people extremely shocked: it didn't look real!

Xu Shenxue listened to these words, and then suddenly thought of a question: "The product is pretty good, but since it has such high technical content, how much does it cost? How much do you plan to sell it for?"

When Xu Shenxue asked about the cost, the head of the robot team beside him showed an embarrassed expression: "As for the cost, it is still relatively high. According to our estimate, the pure hardware cost is about 200,000 yuan. We have not yet determined the selling price!"

Xu Shenxue heard the loophole in this statement: "Pure hardware cost? You didn't include the R&D cost?"

When this was said, the head of the robot team said with embarrassment: "Well, if the R&D costs are included, then there is no way to commercialize it!"

"According to our calculations, our R&D costs alone amounted to more than 2 billion US dollars. If we include the R&D costs of some sensors and various specialized chips developed with the help of some of our sister companies, the R&D costs of the entire project would probably reach more than 5 billion US dollars!"

"If we include all the R&D costs, the product price will be too high, which may be disadvantageous for the initial commercialization of our products!"

When Xu Shenxue heard this... he couldn't help but be speechless.

No wonder they dare not include the R&D cost. The whole project has a R&D budget of more than 5 billion US dollars... If this is included, if the product output is too low, the price of the product will skyrocket!

For a super single product like the S series mobile phones with annual sales of more than 200 million, the research and development funds of several billion dollars are spread out, but the average sales per mobile phone is actually not that much.

However, if the robot starts at several hundred thousand, the sales volume will be too high, and the R&D costs will be very high.

It looks like this robot project will continue to lose money for a long time!
"However, many of our R&D projects are not independent. They are cooperative projects with sister departments and companies. For example, the EYQ3 chip and the ZY23 chip are existing projects of Zhiyun Semiconductor. We are just taking a ride!"

"For other hardware parts, we also contacted other brother departments and cooperated with supply chain manufacturers!"

"The core algorithm was provided by the Institute of Intelligence. It is one of their many algorithm projects. We just saw that this algorithm was quite suitable, so we took it and used it."

"What we really did ourselves was mainly the follow-up AI training. We developed the sensor array for the hand and the integration of the hardware system. This didn't cost much, just over a billion dollars!"

"Strictly speaking, other R&D funds cannot all be counted on us, after all, they also need to use them for other projects."

The person in charge of this robot project also made an excuse...

What's the use of making excuses? Brother departments and suppliers cooperate with you in R&D, spend money to develop products and then sell them to your robot subsidiary for integration. Do you really expect them to sell them to you at the hardware cost price? They must include both R&D costs and profits.

In the end, the same sky-high price!

But it is a fact that the price is too high and it cannot be sold!

At the same time, the R&D funds have already been spent, and the products have to be sold after they are made. Thinking of this, Xu Shenxue said: "As for the product price, you can discuss it with the group's marketing department. I will let Ji Chenghe work on it for you and see what kind of pricing strategy should be formulated."

"But no matter what the price is, the product still needs to be sold. Gradually cultivate the market. Don't be afraid of temporary losses. The important thing is to build up the market. At worst, the payback period can be extended!"

The research and development of new technologies and new products is very expensive, and Xu Shenxue is well aware of this...

Especially for companies like Zhiyun Group that focus on hardware and artificial intelligence, various R&D costs are actually very high.

Otherwise, Zhiyun Group would not spend hundreds of billions of dollars on research and development every year... It is one of the companies with the highest R&D expenditures in the world and is recognized as a big spender.

Compared with Zhiyun Group's speed and scale of burning money, other types of domestic companies are rubbish and can only sigh in envy!
Zhiyun Group's R&D expenditure is approximately US$40 billion, ranking first in the world, and exceeding the total of the second, third, and fourth largest companies!

Because the R&D expenditures of the following companies are basically more than 10 billion US dollars, and the R&D expenditures of the three companies combined are not as high as that of Zhiyun Group alone.

Most of Zhiyun Group's R&D expenditure is spent domestically, with a small portion spent on overseas R&D centers. The domestic portion is around US$6.8 billion, which is approximately billion Chinese yuan at the current exchange rate of .

Zhiyun Group's R&D expenditure in China alone is about 270 billion yuan, while the country's total scientific research expenditure is only more than 1.6 trillion yuan, of which the corporate portion is more than 800 billion yuan... Zhiyun alone accounts for more than 270 billion yuan.

This shows how important Zhiyun Group is at present and how great its driving effect on technology is!
At the same time, this is also the reason why many of Zhiyun Group’s products adopt a high-price, high-profit model... If the product prices are too low and the profits are too low, operations cannot be maintained at all.

Zhiyun Group's business model is actually a very special "high investment and high return" model, which is very different from other enterprises.

Other companies will not spend money as crazily as Zhiyun Group!
But Xu Shenxue dared to do so because he had a scientific research system that could ensure that these huge R&D investments could achieve certain results.

If other companies were to do this, they would have gone bankrupt long ago!
Huge investment in R&D funds, coupled with the assistance of Xu Shenxue's scientific research system, and some foresight brought by Xu Shenxue's previous life.

The above trinity enables Zhiyun Group to continuously develop various new technologies and new products, thereby maintaining the group's huge revenue.

This forms a virtuous cycle of high investment and high returns!
This is something that other companies cannot do, and even many people cannot understand.

In recent years, countless people, whether they are scholars studying economics and business, or managers of other companies, have been working hard to study the Zhiyun model.

They want to know why Zhiyun Group and other companies under Xu Shenxue have been successful?

Why is R&D efficiency so high?
The same R&D talents were just average when they worked in other companies, but after they started working as developers in Zhiyun Group, they all became smart.

But until now, there is still no clear conclusion in academia, and it can only be judged by some vague words.

For example, the working atmosphere, the company's scientific research culture, the perfect organizational structure and other miscellaneous things.

However, other companies actually have these things, especially some other large technology companies. They do very well in terms of corporate culture, office atmosphere, R&D environment, etc. However, their R&D progress cannot compare with Zhiyun Group.

This is very confusing.

Therefore, many people attribute this inexplicable phenomenon to Xu Shenxue's unique ability to lead a team, his outstanding personal scientific and technological vision, as well as his prestige and leadership qualities.

This is also the reason why Xu Shenxue is highly regarded by the outside world, even abroad... He can always lead his team to complete seemingly impossible tasks and produce black-tech products.

In fact, all of these are the result of scientific research systems, some cognitive experiences from previous lives, and a willingness to spend money.

In fact, it has little to do with Xu Shenxue’s personal ability!

His ability to manage a team is actually very average. Most of the time he chooses to let it go... and he even gives it a nice name: "Let professionals take care of professional matters."

Because he really doesn't know how... Even the entire organizational structure of Zhiyun Group and so on, was not created by him, but by the elite talents under him.

After the meeting, most of the elites below, including senior management and a consulting team composed of academics, came up with a solution that was generally accepted.

What else can he say about this plan? Just sign and approve it!
This also makes many senior company managers and academic leaders in the consulting team feel that Xu Shenxue is a very charismatic boss... Who doesn't want to have an obedient boss!
Then these people often boasted about Xu Shenxue's good points in some external public occasions, such as his charismatic personality and strong abilities!

To be honest, apart from occasionally bragging based on his memories of his past life, being willing to spend money, and having a scientific research system to support him, everything else he does is just average... at least that's what he thinks.

However, it is also a fact that Xu Shenxue did create Zhiyun Group.

So in the eyes of many outsiders, Xu Shenxue is shrouded in a halo!
----

No matter what outsiders think, Xu Shenxue himself has some self-knowledge, so he still has the same attitude as before and lets professionals talk about professional matters.

The same is true in the field of robotics, where many achievements have been made.

When Xu Shenxue first demonstrated the concept humanoid robot at a press conference, it caused quite a stir and made many people think that the era of robots was arriving.

And now, the robot project team of Zhiyun Group has developed these two household robots that look very sci-fi.

After looking at these two household robots, Xu Shenxue also took a look at a multi-legged robot, which is often called a robot dog by outsiders.

The usage scenarios of this multi-legged robot are rather special, and one of its main tasks is to perform transportation tasks.

For example, it can be used as a means of transport to accompany personnel and to carry additional supplies.

However, its uses are also limited. After all, it is electric. If you are operating in the field on the battlefield, where can you charge it?

Therefore, it is better to use transport-type drones, because drones can be deployed in the rear logistics base, and the efficiency of drones can be improved by directly replacing batteries after completing the mission.

Therefore, transportation tasks can only be secondary, and the more important thing is to replace personnel to perform more dangerous tasks in some special scenarios.

For example, carry a rocket launcher and fire a rocket at the opponent after passing through a dangerous area!
The technical difficulty of this multi-legged robot dog lies in two points. One is the high-speed maneuverability in the wild, on complex terrains such as building ruins, which requires good passability in complex terrain and also requires maintaining a certain high speed.

The second one just involves target identification and determination...how to distinguish between enemy and friend, armed or civilian threats.

The third, also very important and very complicated issue is whether to give it independent decision-making power.

In terms of autonomous decision-making power, the attitude of the cooperative organizations is still quite conservative. They adopt a remote control approach to decision-making power and do not give any unmanned equipment the right to fire for the time being.

Whether it is a drone, a robot dog, or even an AI fighter jet, etc., the right to open fire is not given for the time being, and the right to open fire can only be issued through remote control commands.

Therefore, the autonomy of all current unmanned equipment research and development is limited to other areas such as action and information collection, while the right to open fire is strictly limited to the level of remote control command.

After all, no one is brave enough to give AI the right to fire... If AI makes mistakes in identification and decision-making, it will be a very serious problem.

Because the right to open fire is the absolute bottom line.

As for the many issues mentioned above, except for the right to open fire, there are actually no rigid technical restrictions on many other technologies for Zhiyun Group at present.

Now Xu Shenxue has seen several multi-legged robots in this laboratory, including traditional four-legged robots and six-legged robots.

There is also a wheeled multi-legged robot, which can use wheels to maneuver as quickly as a wheeled robot, while also having the complex terrain passability of a multi-legged robot.

During the test demonstration, it showed very good rapid maneuverability.

"These multi-legged robots are modular unmanned mobile platforms developed by us, while the identification and strike modules were developed by our partner organizations. We are responsible for the development and supply of the mobile platforms, while they develop the identification and strike modules themselves."

"According to the plan, after the two are combined, they can be combined into multi-legged robots with different purposes according to different modules to perform different tasks."

"In August, our four-legged mobile platform was combined with an enemy strike module to create a multi-legged robot for complex urban terrain. The effect was quite good. It can conduct reconnaissance and strike inside buildings."

"We are also trying to combine it with a six-legged mobile platform. Our six-legged platform has better carrying capacity, platform stability and passability, and is more suitable for complex terrain, especially maneuvering in wild terrain!"

Xu Shenxue looked at a four-legged robot dog in front of him, with a model of a gun on its back. Its appearance was actually not much different from the robot dog he first saw in the laboratory a few years ago.

Back then, he took the robot dog and drone from his laboratory, as well as a small host equipped with AI, and went to Beijing for a meeting. In the end, he won a lot of support in the semiconductor field!

Now, several years have passed, and the robot dogs and drones that once only existed in laboratories and required a special host to use have gradually matured.

Xu Shenxue said: "This is also good, let's continue. There are not many of these multi-legged robots in the civilian market, so in the subsequent specific research and development, we still need to listen to the ideas of our partners, strive to quickly form them, and then launch them to the market as soon as possible!"

After reading about the multi-legged robot project, Xu Shenxue left the robot project team and then called Ji Chenghe to discuss with him the issue of bringing this domestic robot to market.

There are so many robot research and development projects within the group. At present, many seemingly good industrial robots and service robots do not have much market: robots cannot compete with humans!

After all, industrial robots are sold to companies, and companies only look at costs in order to make money, and have no intention of paying a premium for black technology or sci-fi products...

Therefore, for a company like Zhiyun Group, it would rather make consumer products for individual users than industrial products for enterprises...corporate customers cannot offer a premium!

Even though multi-legged robots are in demand in special fields, they are just that, and are not easy to export. Don’t expect to export such advanced and sensitive things in the short term... So it’s not easy to make money from products in special fields!
Money is not easy to come by in industry and special fields, which is why Zhiyun Group has spent a lot of money on the research and development of household robots in the field of robotics... They can count on a large group of individual consumers to pay a premium for black technology, science fiction, and new experiences.

Although domestic robots theoretically face competition from human domestic workers, in reality the two are not competitors.

Because no one would really regard it as a pure household robot.

For some people who don't have hundreds of thousands of dollars to spend, they buy this thing to play with and experience the black technology that can only be seen in science fiction novels and movies.

When the first generation of the Hailan SEC was first launched on the market, it was also extremely expensive and far from being practical, but it still sold very well... It relied on the sci-fi feel brought by black technology. Those rich people bought the Hailan SEC just to have fun and never expected this car to be used as a daily car!
Today’s household robots are actually following this path. They are nominally household robots, but in fact, Zhiyun Group has never expected consumers to buy them to do housework…

No one would really buy this thing to use for housework, right?
In fact, in Xu Shenxue's mind, the positioning of this product is: a large toy!
Just a little expensive!
Now that the household robot has been developed, the technical personnel have completed most of their tasks, and the pressure has shifted to the marketing team headed by Ji Chenghe.

How can you sell such a big toy whose hardware cost is over 200,000 yuan?
(End of this chapter)