Rebirth 08: Rise from copycat phones

Chapter 364 Black Technology: Housekeeping Robot
Since October, many places in the north have seen a significant drop in temperature, entering the autumn mode with late nights and cool water.

But in Shenzhen in the south, October is still as hot as a furnace. People on the streets are wearing thin clothes. If you stay outside for a while when the sun is at its strongest at noon, you will sweat all over.

In this hot autumn, most of the office buildings in Shenzhen Headquarters Park are as bustling as usual.

As the largest scientific research base in China and even in the world with the largest number of employees, Zhiyun Shenzhen Headquarters Park has become a global focus since the day it was born.

Here, the world's top artificial intelligence R&D base, the world's top smart terminal R&D base, and the world's top chip design base are gathered together.

It is also the place where a lot of innovative technologies are born!
According to information released by Zhiyun Group, after two expansions, the Shenzhen Zhiyun headquarters park now has more than 80,000 people working at the headquarters base, of which more than 90% are technical personnel, and other non-technical positions such as administration, legal affairs, and finance account for only about 10%.

In such a headquarters base, a large number of new technologies and new products at the forefront of current science and technology are gathered.

The robot project is one of them!

The 16th, 17th and 18th floors of the office building D in the headquarters park are the offices of Zhiyun Robotics Co., Ltd., a subsidiary of Zhiyun Group.

Zhiyun Robotics Co., Ltd. is also a subsidiary of Zhiyun Group responsible for the development of robotics business.

At this time, in the office building on the 16th floor, the management of Zhiyun Robotics was reporting to Xu Shenxue on the group's robot business and R&D.

A middle-aged bald man introduced: "Our robot project team has been established for six years. We have launched industrial robot projects, catering robot projects, housekeeping robot projects, police robot projects, and military robot projects!"

"Among them, the industrial robot project is our earliest project to be developed and commercialized. We have launched more than 20 different types of industrial robots to the market, which can replace manual labor and traditional robotic arms for industrial production. We have provided products to more than 100 customers around the world for in-factory production."

"Our industrial robots are actually a type of robotic arm, but they are highly intelligent and can learn and evolve autonomously. They can complete more complex workflows. Compared with traditional robotic arms driven by fixed programs, they have greater flexibility and wider adaptability!"

"The demonstration dark factory currently built by Weiku Industry has adopted a large number of industrial robots from our project team, and the effect is very good!"

"It's just that... the cost of our industrial robots is still relatively high. Apart from a small number of technical verification demonstrations in China, they are not of practical use value for the time being. The main market for our industrial robots is overseas where labor costs are relatively high, especially in Europe. We have already provided dark factory solutions to many European companies!"

"The catering service robot series is our second commercial product launched on the market. Currently, there are three main series: beverage making series, light meal making series and catering wheeled transportation series."

"For the beverage making robot, we have completed the production algorithms of hundreds of different beverages, and can provide customized services according to customer needs. At the same time, customers can also customize the beverages to ensure that the beverages made perfectly meet customer needs!"

"After it was officially launched on the market last year, it has already opened up the market in some areas with relatively high labor costs. After all, although our robots have a relatively high one-time cost, there is no need for subsequent wages or other additional expenses. It can be just regarded as an ordinary equipment investment, which has a very large cost advantage in areas with relatively high labor costs."

“In order to better serve our customers, we will continue to train and develop the production of new categories of beverages and foods, and provide customers with upgrade packages in the future!”

"Our marketing department believes that after about a year of market cultivation, our catering service robots are likely to see a significant increase next year!"

"Of course, this is only limited to areas with high labor costs. In places where labor costs are too low, such as a large number of developing countries, our robots still do not have a competitive advantage. This is why the main markets for our industrial robots and catering series robots are overseas!"

"As for this side of the country...Alas!"

The head of the robot project team continued to introduce:

"Industrial robots and catering robots are our official commercial products. Although the market is very niche, we are the only player in the market, so our overall revenue is barely acceptable. These two products are expected to bring in about $1.2 billion in revenue this year!"

Xu Shenxue nodded slightly after hearing this: "Not bad. It is very rare to be able to develop products and commercialize them to generate revenue in the early stages of the development of the robotics field!"

The project leader continued: "In addition to the catering service robots, we plan to officially release our new products next month: two housekeeping robots!"

"There are two types of this robot, a bipedal, two-armed anthropomorphic form, and a wheeled, two-armed form!"

“Although the humanoid bipedal robot is still too complicated due to its power mode and insufficient endurance, this product is more in line with human’s understanding of robots. So even if it is not very practical, we still choose to make it and prepare to put it on the market!”

"Of course, the real product we are preparing is still the wheeled dual-arm housekeeping robot!"

Xu Shenxue said: "This is quite interesting, but your training progress is so fast? I remember that your housekeeping robot said it would take three years to complete the training and mature?"

At this time, the other party said: "It is like this. According to our estimation, if we rely on ourselves to complete the training, it may take more than three years."

“This takes too long and costs too much!”

"And if only our employees are used for training, the data samples are too small, and the products trained after three years will not be perfect!"

“So we decided to take another path and let users do the training for us!”

“When we release it, we will demonstrate the robot’s autonomous learning capabilities and require data upload and usage permission. We will allow a large number of users to become our trainers and help us quickly accumulate a large amount of data!”

Xu Shenxue almost wanted to laugh when he heard this... These people were too smart to actually ask users to help with the training, and they even paid for it...

But if you put yourself in the shoes of a R&D staff, then it will seem natural!

Whether it is a household robot or any other type of robot, its core technology is not the robot, but artificial intelligence!

The algorithm inside is the most core and most critical technology!
Of course, sensors that collect data and computing chips that support algorithm operation are also important.

The characteristics of artificial intelligence determine that the more data, the better. When Hailan Automobile was developing autonomous driving, it installed a large number of sensors in the car, cooperated with a large number of taxis and online ride-hailing companies, and continued to collect a large amount of driving data... Why can Zhiyun's driverless car project produce such a good autonomous driving?

These large amounts of driving data are a very important part of this.

Later, Hailan Automobile continued to collect data in its own cars, so it improved itself faster;
After all, artificial intelligence is trained!

Although Zhiyun Group's artificial intelligence technology is already quite good, it is still a type of AI that requires targeted training.

If you want it to do any kind of work, you must conduct long-term, professional data feeding training... and you have to train it one type at a time.

This kind of training is actually very time-consuming.

This is also the reason why Zhiyun Robotics Co., Ltd. will first launch a relatively simple series of catering service robots and industrial robots for fixed occasions to the market... the training volume is relatively small.

Because Zhiyun Robotics Co., Ltd. has only trained so many functions at present... Later, we have to train one after another, such as sweeping and mopping the floor, tidying up the living room, tidying up the kitchen, etc...

Many of these housework that are not difficult for humans are actually very difficult tasks for robots.

For example, cleaning the kitchen... From a human perspective, it's nothing more than washing the dirty dishes, pots, etc., and at most mopping the floor.

However, from the perspective of robots, it can actually be divided into dozens of different jobs, any one of which is highly difficult and requires special training.

This kind of training is very time-consuming... Unlike text, voice human-computer interaction has a large amount of mature and available data that can be directly fed in.

It is not like driverless cars, which have a large amount of vehicle driving data that can be used for training.

The data needed for the research and development of a large number of robots is actually difficult to collect on a large scale, and we can only accumulate data slowly and conduct training.

In the current field of artificial intelligence, the core algorithm framework is the most important, followed by the computing power for training AI, and then the data fed into it.

The above three are a trinity!

The first two are limited by hard technical conditions, while the latter is purely a matter of money and time.

In robot development, the biggest time cost is not the development of the algorithm itself, but the feeding and training of large amounts of data... For example, if you want a robot to clean the table, you have to constantly correct the robot's movements so that it knows how to clean it.

Therefore, the seemingly high-end robot development... in addition to the various software and hardware development in the traditional sense, there are more ordinary people who constantly correct the robot's mistakes.

Of course, this low-tech, repetitive and boring work is not carried out in a high-end headquarters base, but in a training factory in the suburbs... There, at the training base of Zhiyun Robot Co., Ltd., there are more than two thousand employees who continuously train and correct robots.

Let the robot learn to wash dishes, wash clothes, sweep the floor, mop the floor, pour water, make coffee and other actions that people are accustomed to in their daily lives.

It is hard for ordinary people to imagine that such a high-end robot was trained by a group of ordinary workers with low education, such as junior high school students and high school students...

The result of their previous training is a series of catering service robots... which are already quite complete and mature. The robots can skillfully make coffee, milk tea and some other simple foods according to instructions.

It can be used in simple, light-food catering industries such as coffee shops, milk tea shops and even western restaurants, allowing investors to save a lot of labor costs.

However, after completing one housework task after another and accumulating experience, the robot project team gradually completed the research and development of the housekeeping robot!

And this thing, for contemporary technology, is undoubtedly black technology!

Even though it’s just a stupid-looking household robot and isn’t very good at what it does, it’s still veritable black technology!

Xu Shenxue is very much looking forward to this and plans to go to the scene to see the real thing soon.

But now the meeting has to continue, and the current robot project team continued: "In addition, we also officially launched an industrial transport wheeled robot, or an industrial transport unmanned vehicle, to the market at the beginning of this year. This product was developed in cooperation with the unmanned vehicle project team!"

"The product sizes range from as small as one meter to as large as ten meters. Currently, our industrial unmanned vehicles have been used by some customers!"

“The main application scenarios are in storage centers, logistics centers, and ports!”

"Last month, we reached an agreement with a terminal operator in the United States to provide the company with a complete set of unmanned terminal transportation solutions for container transshipment at port terminals. According to our estimates, after the implementation of the solution, it will save our customers more than 40% of costs and improve terminal operation efficiency by more than 30%."

"Industrial transport unmanned vehicles are the main field of our company's future and the field most likely to bring in a lot of revenue for the company in the near future!"

"Of course, what we hope for most this year and even next year is to make a breakthrough in domestic robots. After the first generation of products are launched, they will be liked by some technology enthusiasts, and then they will train us, so that we can obtain enough data!"

"As long as there is enough data, we will be able to launch a relatively complete domestic service robot next year, a real domestic robot, and even extend it to other fields, such as elderly care, child care, etc.!"

"The market for domestic robots is extremely vast. Once fully developed, it will be a trillion-dollar market."

Xu Shenxue said: "The market for domestic robots is naturally very broad, the premise is to make good products. Come on, the meeting is almost over, take me to see the real thing, good things must be brought out for inspection!"

(End of this chapter)