Great Power Heavy Industry: Rise

Chapter 190 Brainstorm
The second day of the daily three-phase project kick-off meeting.

Zhong Cheng held a seminar on the design of intelligent robots. In addition to calling all the project members in charge, he also invited the main backbones of other project teams.

It is now approaching the end of 2018, and the deadline for the completion of the task that Zhong Cheng promised to the country is not long.

Although the leaders did not issue mandatory requirements, he still set goals for himself.

At the beginning of the meeting, Zhong Chengcheng asked Zhao Wei to explain the problems existing in the current research and development of intelligent robots.

After listening to Zhao Wei's explanation, the participants discussed a lot.

Most of them are responsible for their own projects, not aware that intelligent robots still have such a big problem.

Zhong Cheng had a headache. Zhao Wei was not very competent as the project leader, and the communication work was not done well.

Are you considering making Fan Chengjun more responsible for communication and coordination?

Let's talk about this matter later.

He let everyone speak freely and say what comes to mind, it can be regarded as a brain storm!

Everyone chatted in unison, and all kinds of wonderful ideas were put forward.

Most of them are unreliable. Zhou Hu sorted out the more valuable parts, and everyone discussed them one by one.

Jiang Yan proposed to subdivide the functions of intelligent robots.

Military robots, industrial robots, domestic robots, service robots, etc., each of which is subdivided.

For example, industrial robots are divided into handling robots, welding robots, assembly robots, mining robots, etc. according to their functions.

Subdivided in this way, the requirements for each professional robot are not high.

But in this way, it will be too difficult for future robot manufacturing factories. How many robots with different shapes and functions will be produced?
This plan was rejected.

Gui Bin's assistant, Huang Xueshen, also proposed an idea.

Establish a central control server, each robot is just a terminal, and the complex storage and computing functions are handed over to the central control server group for processing.

This does not require such a complex control system for each robot.

However, this solution still cannot solve the problem of robot movement, and there is a risk of delay, lag or even interruption in the information interaction between the server and the terminal.

This proposal was also rejected.

The next few ideas were also rejected on the spot.

The meeting room fell silent, and dozens of people were frowning.

Just when Zhong Cheng wanted to end the meeting and discuss it another day.

Gui Bin raised his hand to ask for the floor. He said he had a new idea.

Zhou Hu made a joke, asking if he wanted to use quantum computers to solve this problem.

Gui Bin was very embarrassed. He recently proposed to develop a quantum computer, but was rejected by Zhong Cheng, thinking that the time was not ripe, so don't waste time.

Zhong Cheng glared at Zhou Hu and asked Gui Bin to express his thoughts.

Gui Bin stood up and said that his ideas were not yet mature, and he could only be considered as an introduction.

He said that the current optical chip processor can be made very small. A micro-optical computer can be as small as the size of a fingernail, and the cost has also been reduced to a few hundred yuan.

Is it possible to install a large number of miniature optical brains in the body of an intelligent robot to form different functional modules and handle different affairs respectively.

For example, for body control, a dedicated sports management module can be established. This module can also be subdivided into sub-modules such as stable standing, walking, and running, which are controlled by different micro-optical brains.

Instructions are issued from the central computer of the robot, and the motion management module decomposes the instructions, and then sends them to each sub-module, or even a lower-level module.

In this way, very complex movement coordination is performed by hundreds or even more micro-optical brains, and the control system in each micro-optical brain is relatively simple.

In the same way, the processing of external information can also use this method to establish different functional modules and decompose the difficulty.

Zhao Wei was a little dumbfounded by Gui Bin's thoughts, "How many miniature optical brains will be used, and such an intelligent robot will not be a sky-high price?"

Gui Bin explained: "You may not be aware of the development of optical chips. Now that optical chips have entered mass production, various specialized manufacturing equipment has begun to be used, and manufacturing costs have been falling sharply."

"If a large number of intelligent robots are launched, the production capacity of optical chips will be further expanded, and the cost will drop to tens of yuan, or even a few yuan."

"Then even if tens of thousands of optical chip processors are used in an intelligent robot, the cost is within an acceptable range."

Indeed, if the cost of an intelligent robot is several hundred thousand yuan, then there is the possibility of large-scale promotion and application.

This idea does work!

"Clap clap clap!"

Zhong Cheng took the lead in applauding Gui Bin, Zhao Wei followed suit, and the conference room was full of applause.

"Gui Bin, I really have you, you have a flexible mind!"

Zhao Wei praised Gui Bin without hesitation, he was the most excited to be able to solve this problem.

Gui Bin sat down with a flushed face. He has always had a weak sense of presence in the team, and this time it was a sigh of relief.

Zhong Cheng found that Gui Bin had made rapid progress over the years and had become a real expert in the field of computer hardware.

Building a super-large optical brain network in a robot is not something ordinary people dare to think about.

With Gui Bin's whimsical idea leading the way, other people's brain holes were also opened.

New ideas are constantly coming up that complement the smart design solutions.

Li Mei's assistant, Zhang Liang, has always been down-to-earth, unobtrusive, and is very important to Li Mei.

This time, Zhang Liang put forward a very creative idea, which can be regarded as a complete solution to the computing problem of intelligent robots.

He proposed a brand-new solution for intelligent robots to process external information.

His idea is based on the fact that more than 99% of humans are actually very limited in their range of motion.

The place of residence, the place of work, the limited places of consumption and entertainment, together with the connecting routes among them, constitute the daily activity space of most human beings.

If the range that the human eye can observe is taken as a scene, the information received by other sense organs is also basically limited to this scene.

After careful calculation, the number of scenes contained in the areas where human beings are often active will not exceed 1, or even less.

For example, human dwellings can be covered with more than a dozen scenes.

And there are even fewer active scenes for intelligent robots. When they move in these limited scenes, they don't need to record all the scene information every moment.

They only record information about scene changes, and the changes that their sensors can recognize or allow to recognize are actually very limited.

In this way, the external information they need to store and process every day should be less than one ten-thousandth of the real-time record volume.

Because the basic models of these scenes are pre-installed on their optical brains or when they enter for the first time, what needs to be stored and processed is only the amount of scene changes.

These scene models can be shared through networking, and updated data can also be shared.

More importantly, with the continuous improvement of thousands of intelligent robots, these scene models can even build a completely real human social model in the Super Fuxi Optical Brain.

This method can also be applied to the learning of intelligent robots.

The accumulation of every bit of knowledge and experience of each robot in daily work, through sharing, will bring progress to the entire robot system.

Such progress is magnified by millions of times, which is very amazing.

To put it simply, if a robot has learned it, all robots will know it. This is the advantage of computer arithmetic.

Zhang Liang's speech shocked everyone.

If Gui Bin builds a large-scale network inside the robot, then Zhang Liang connects all the robots into a super-large network.

Zhang Liang's idea not only solves the calculation problem of intelligent robots, but also solves the learning problem of intelligent robots.

The "intelligence" problem of intelligent robots is perfectly solved.

(End of this chapter)