Global Times Mobile

Participants interact with a robot at the 18th China High-Tech Fair in Shenzhen, South China's Guangdong Province in November 2016. Photo: CFP

The artificial intelligence (AI) revolution is under way in China, with its applications gradually entering more and more aspects of people's daily lives.

There are a wide variety of scenarios in which AI technologies can be applied in China: In hospitals, AI-driven models can help diagnose diseases by consolidating a database of information; in factories, robots equipped with visual tracking and sensor systems can help workers move components around on a conveyer belt, as well as identify different items and carry out pick-and-place work with precision. In banks, customers can ask questions to a humanoid robot about what kinds of services they need.

The sector is expected to grow even more rapidly in the next three to eight years, as the central government has shown its determination to help AI-related companies grow and make technological breakthroughs, according to industry representatives. 

By 2020, the AI industry will become a new economic growth engine, according to a guideline unveiled by the State Council, China's cabinet, on Friday. The market scale will surpass 150 billion yuan ($22.2 billion), and advances are expected in major technologies including modeling, core components, high-end equipment and basic software, the guideline showed.

By 2030, China's AI technologies should be on par with the best in the world, and the country should become an innovative hub for AI technologies. The value of the sector and related industries should be in the range of 10 trillion yuan, said the guideline.

"The guideline will help the AI sector to grow in a sustainable way. However, it should clarify more which key technologies should be targeted and which sectors should further develop them," Lu Yanxia, an industry analyst with IDC, told the Global Times on Monday.

Top-level design for the AI industry will bring more capital as well as support from the government to the sector, which will initially help the application of AI technologies in government-related business to grow, Lu noted.

Eight key technologies have been highlighted in the AI industry guideline, including knowledge computing, reasoning, close interaction between machines and human beings, self-driving technologies, virtual reality and modeling, narrative language processing and machine learning.

Global hub

China is becoming a hub for global AI development, and the technology has now moved beyond the lab, with many machine-learning systems already in commercial use for a wide variety of applications, according to a report released by the McKinsey Global Institute in April.

"A robot, which is one of the containers of AI, should be considered as a species. Although a strong-AI or a super-AI are unlikely to suddenly happen, a robot can have the ability to think, step by step," Jiang Huabing, CEO of Shanghai Clever mRobot Technologies Co, told the Global Times on July 18.

The company, which has developed a service robot called Cooky, can carry out sentiment analysis by scanning a person's facial expression. When it is placed in a bank or an administrative office, Cooky can respond to people's requests based on scenario learning modules.

Jia, the former founder of the smartphone business at Huawei Technologies Co, has applied the business model favored by his former employer to the robotic start-up. By working with different software and hardware providers, mRobot is in charge of assembling different components and improving user experience.

However, it had to spend a large amount of money on buying core technologies such as voice recognition from other software developers iFLYTEK and Xiaoi.

Challenges remain

In the technological aspect of China's AI revolution, challenges still remain, Lu noted. "For example, speech recognition, as part of AI technology, is still in the early stages of development, and does not include semantic understanding or analysis."

However, in image processing, the environment where the technology is applied has very few obstacles, which "is still easy to handle," the analyst noted.

In a workshop at Shanghai-FANUC's factories, a robotic arm equipped with a visual tracking system is being tested.

"Although we can use image processing to carry out item-selection work, the question of machine learning is difficult to resolve," said Qian Hui, general manager of the company.

Machine learning is one type of AI technology that Shanghai-FANUC is focused on, but it is also receiving a great deal of attention by giant tech leaders such as Microsoft and Google.

The world was amazed by Google DeepMind's AI program, AlphaGo, which used machine learning to defeat its human rival in the board game Go. But the research giant has also come up with open-source software, which has become a machine-learning toolbox available for scientists, researchers and CEOs, media reported in June 2016.

Some tech companies have been increasingly investing in research and development with the aim of coming up with self-developed modules to better analyze collected data.

For instance, Chinese online car-services provider Didi Chuxing can optimize transportation capacity through data mining, machine learning and cloud computing in order to maximize the efficiency of transportation systems and improve people's travel experiences, according to a statement Didi sent to the Global Times on Tuesday.

Great expectations

The country's vast population and diverse industrial mix can generate huge volumes of data and create an enormous market, according to the McKinsey report.

While some companies are focusing on making technological breakthroughs, others see their opportunities in exploring a wide range of scenarios in which AI technologies can be applied in the near future.

"Now that the central government has shown its support for this industry, AI technologies can first be used in smart cities, financial institutions and services," said Lu, the analyst.

In terms of the application of AI technologies, the scenario for speech processing is considered the most mature one for now, but will see its growth slowing down, according to a report published by CCIC in June.

In addition, image processing technology can easily be used in many sectors, including security, hospitals, marketing and driver assistant systems, the report noted.

AI is not a fairy-tale anymore and requires the further development of AI and applications, iFLYTEK said in a statement sent to the Global Times on Monday, adding that its goal is to meet consumer demand.