The top female star who can make a lot of money for the country has 70,000 fans a day, all because of this AI?

What makes the top panda world receive 70,000 fans a day? What makes a factory's daily output exceed one million? This is all thanks to the AI ​​behind them.

Recently, the video of this top female star has hit the Internet.

At 8:30 in the morning, the Chengdu Research Base of Giant Panda Breeding was crowded with people, all coming to see how Huahua was snatched away from the bamboo shoots.

There were so many onlookers that the security guard was helpless, "There are so many giant pandas here, do you only want to see this one? You love it too much."

# Speaking of which, it’s all about seeing pandas, but the treatment of tourists in China and abroad is very different.

It is very difficult for fans to meet the extremely popular Wang Xiangxiang from Ueno Zoo in Tokyo. Before Xiangxiang returned to China, 60,000 people across Japan signed up to see her one last time. The winning rate was about 1:24, and the winning rate for the last group reached 1:70.

But in China, we have fully realized the "freedom to watch pandas".

The Chengdu Giant Panda Breeding Base where Huahua is located had a total of 72,000 visitors per day at its peak.

With such a high flow of people, how to deal with security, ticketing, and on-site order?

At the 2023 Megvii Enterprise Business Partner Conference, Yang Yong, founder of Hesheng Zhilian, mentioned when sharing that one of the keys is the transformation of a three-code-in-one system including the ticketing system. This The system is jointly completed by Hesheng Zhilian and Megvii.

Behind the unexpected silky smoothness

In fact, there are many scenes that may not be noticed at ordinary times, but are very "silky" to experience, all with the blessing of AI.

For example, behind the toys we play with and the snacks we eat, there are actually a group of robots working hard to "move bricks."

In the Hsu Fuji Factory in Dongguan, there is a group of robots that carry candy and sing songs, and they do the work.

The candies on the assembly line are thrown to the robot by the mechanical arm, and then they are moved onto the driverless truck and transported to the warehouse. They are all done in one go, so silky smooth.

These small AGV handling robots can also recognize the way and plan their own routes. Just scan the QR code on the ground to know where they should go.

Look, although the little robot is small, it is very strong. It can easily carry cargo and really carry everyone.

There are 25 of these powerful little robots in total, and they all accept the command of the Megvii Hetu system. The system automatically receives orders, automatically schedules, and automatically plans, just like that.

After being taken over by AI, the site does not require manual care and management at all. In the past, raw materials, semi-finished products, finished products and other materials were transported manually or by forklifts. Now, the timeliness and accuracy of material handling have reached 100%, and the robots can automatically avoid obstacles, greatly reducing workshop safety risks.

After the AI ​​algorithm robot was put on the job, Xu Fuji’s labor costs were greatly reduced. With production capacity remaining flat, the number of employees during off-peak and peak seasons has been reduced by about 40%. The 23 people in a certain warehouse buffer zone have been completely replaced by 25 AGV robots, saving 1.8 million yuan a year.

However, AI does not simply eliminate humans. Xu Fuji’s current job requirements are more professional, such as dual talents in IT and food industry.

So, we humans still have a chance (dog head).

A toy factory’s self-rescue

Actually, the biggest problem facing factories now is that they can’t recruit people.

Young people don’t like to work in factories, and many factories are facing labor shortages. If there are not enough manpower, no matter how many orders there are, we will not be able to fulfill them.

Shiwanxin Toy Factory is facing such a problem. After the Spring Festival of 2020, colleagues who went to Fuyang, Anhui Province to recruit workers, found no one.

At that time, although the factory was partially automated, a large number of workers were still needed in various processes of the production line, transportation of items and ingredients after the production line, warehouse inventory, and outbound transportation.

As a last resort, factory owner Lin Guanjun recruited at high prices through his connections and asked existing workers to come back early to catch up on work, and finally passed the crisis. It was at this time that Lin Guanjun made up his mind to carry out full automation transformation.

This decision is actually very difficult. The previous semi-automatic transformation has burned nearly 20 million in four to five years.

What equipment should be used for full automation? Who do manufacturers look for? How much does it cost?

After hesitating for three months, when he visited the well-known Bull Intelligent Warehouse Renovation Project in Ningbo, Lin Guanjun's eyes lit up with a few words from the workers: "This is our second phase of the warehouse project in cooperation with Megvii. Correspondingly. Compared with the first phase, our production capacity has tripled, from 5 billion to 15 billion."

After cooperating with Megvii, Shiwanxin's smart three-dimensional warehouse was successfully completed.

Pallet stackers, container stackers (Miniload), box storage robots (CTU), large and small AGVs (or "AMRs"), conveyor lines, and elevators operate in an orderly manner. The AMR robot car developed by Megvii shuttles through it, connecting the production workshop and the central warehouse, realizing the linkage between the warehouse and all production lines.

Ershiwanxin has also become the benchmark for local smart factories in Taizhou. Megvii’s 3A smart logistics solution (AS/RS AMR AI) can coordinately dispatch 8 different types of automation equipment through Megvii Hetu.

Today, Shiwanxin is expected to produce 20 million sets of wooden toys annually, with a daily output value of 1 million yuan. The utilization rate of warehousing space has increased by three times, and the production efficiency has increased by more than 2 times, it also greatly reduces the manual labor intensity, and the manual links are reduced by 60% under the same production capacity.

Why the physical industry needs AI

So, why have these factories chosen the path of AI intelligent transformation?

Take Shiwanxin as an example. Although automation costs a lot, it also improves product yield, which in turn drives revenue growth. And its intelligent self-rescue also allows us to see the huge role of AI technology in promoting the real economy.

In addition, Sinopharm Guangzhou’s logistics center can save tens of millions of yuan in the next five years after undergoing digital intelligence transformation; Aixu, a leading new energy company, has achieved this through 140 Megvii SLAM robots. With automatic material handling and high-precision docking, the daily production capacity of battery silicon wafers exceeds 3 million pieces.

It is not difficult to see from these actual cases that the deep integration of artificial intelligence and the real economy has become the development trend of the industry. So in summary, it can be summarized into three points:

• Digital technology and physical industries are accelerating the integration

For physical enterprises, the use of robots, AI, IoT, etc. Automation and digital intelligence upgrade of technology are inevitable choices to reduce costs, increase efficiency, optimize management, and improve employee experience.

• The real industry shifts from demographic dividend to technological dividend

Currently, the labor shortage problem is becoming increasingly prominent, and the digital labor force created by AI technology can not only fill positions, but also follow people. Form a good collaboration.

For example, in dusty or other unhealthy environments, robots can be used for transportation and storage, and employees only need to supervise in the central control room.

• The trend of AIoT software and hardware integration is obvious

In the logistics scenario, AI, which used to be more focused on software, has begun to be combined with various hardware to further become a logistics robot. , and a set of intelligent logistics equipment. Flexible logistics, which is related to overall operating efficiency, also relies more on AI in the scheduling of clustered intelligent equipment.

Secondly, this year’s Two Sessions mentioned that in the context of becoming a manufacturing power and creating new advantages in the digital economy, artificial intelligence companies need to shoulder the responsibility of using AI to empower the real economy, achieve cost reduction and increase efficiency, and better promote The task of high-quality development of society and industry.

It can be said that the integration of data and reality based on AI is in line with the country’s overall direction of accelerating the construction of a modern industrial system and promoting the development of high-end, intelligent, and green manufacturing.

What Megvii has done

As we have seen before, Megvii has helped many physical companies to digitally upgrade their factories by integrating AI technology with automation technology. Currently, there are more than 200 implementation cases in the field of smart logistics alone.

In today's fierce international competition, if you want to open up and develop new fields and new tracks, or even create your own new advantages, you need to rely on technological innovation. The scientific research team of more than 500 people allows Megvii to output complete technical capabilities from the underlying deep learning framework to algorithm development and integrated software and hardware products.

To this end, after years of independent innovation, Megvii has created the AI ​​productivity platform Brain and the algorithm production platform AIS (AI Service).

Brain integrates algorithms, computing power and data capabilities to fundamentally improve the efficiency of AI; while AIS further standardizes the entire process of algorithm production, Not only has the production efficiency increased by more than 50% on average, but the accuracy of the model output also far exceeds the industry average.

It is understood that 60% of the algorithm products generated in 2022 will be completed using AIS. This year, Megvii plans to provide its algorithm mass production capabilities to partners through the Megvii Rubik's Cube intelligent analysis box, making writing a valuable AI algorithm as easy as making software, or even easier.

In this regard, Megvii co-founder and CEO Yin Qi said: "We are very willing to take our own lives so that our programmer friends can do something more meaningful."

However, if you want to innovate in application scenarios that are already very mature or have not had breakthroughs for a long time, on the one hand, you need partners who are deeply involved in the industry to fully understand the scenarios, and on the other hand, you need the excellent technology of AI companies Support and efficient algorithm development.

For example, in the face of the increasingly diversified, personalized and refined physical education, Jinshan Top has innovatively applied big data and computer vision in physical fitness assessment after fully investigating the key needs of sports reform. Combine.

With the support of Megvii’s core technology, after 180 days of cooperation and research, the two parties successfully launched a physical training product that can realize 10 types of application scenarios, and captured 60% of the market segments in one fell swoop share.

Another example, a coal washing plant of a mining company has always hoped to shorten workers’ working hours and reduce their labor intensity, especially to minimize the need to go down the mine and ensure that workers can work in various situations. ensure the safety of every link and enhance their sense of gain and happiness at work.

After in-depth research, Megvii and Riyadh Electric developed personnel safety and security algorithms such as fall detection, smoking detection, and labor protection equipment wearing identification, as well as belt over-limit particle size monitoring, coal slurry water spillage monitoring, Intelligent detection algorithms for equipment such as bucket grate stacking detection, demediation screen plate shedding monitoring, video linkage alarm, etc.

On the AIoT path chosen by Megvii, companies not only need to focus on technological research, but also on how to turn scientific research into good products.

Megvii believes that AIoT products must lower the threshold low enough in terms of ease of use, certainty, and difficulty in integration and re-development.

Because, if you implement a single point in a certain industry, the process from 0 to 1 has been completed. So what to do from 1 to 10? Lowering the threshold and returning technology to the user side is the basic point for the market to explode from 1 to 10. This is also the development direction of Megvii in the next few years. ​

##

The above is the detailed content of The top female star who can make a lot of money for the country has 70,000 fans a day, all because of this AI?. For more information, please follow other related articles on the PHP Chinese website!