The turning point of the robot industry is coming to break through the focus on perceived intelligence

The robot industry is hot in the world. In March of this year, Alpha Go and Li Shishi's century war marked the beginning of the new wave of artificial intelligence.

To what extent has the robot industry developed so far, and what problems exist and how will it develop? In this interview, researchers from the Hong Kong University of Science and Technology and the Chinese robotics industry were interviewed.

"Now robots have become a strategic high-tech that countries all over the world are competing for, because robots will change our manufacturing model and lifestyle." Tan Jia, dean of the Hong Kong University of Science and Technology, said to the news, "Big data and labor." In the future, intelligence must be integrated in all walks of life, and drive the development of the industry. We are also very happy to see that the big data products represented by Tencent WeChat and Dajiang UAV are far ahead of Hong Kong, Europe and America. Some Western technology companies, even some giants, are looking for ideas from Chinese companies."

In the second half of 2006, Wang Wei, who is also a graduate student at the Hong Kong University of Science and Technology, established Dajiang Innovation in Shenzhen. At present, Dajiang Innovation has become the world's largest manufacturer of consumer-grade drones, selling thousands of sales per year for about 1,000. Dollar quadcopter. DJI Innovation is also the first Chinese brand to explore the world of emerging UAV consumer goods. Chris Anderson, founder of DJ's innovative competitor 3D RoboTIcs, commented on Xinjiang Innovation, "They are like Apple iOS in the drone field."

In Tan Jiayin's view, artificial intelligence will bring a new intelligent revolution to mankind, just like the industrial revolution of the 19th century. In the future, AI will lower the threshold and truly serve everyone, not just a few people and a few institutions.

Despite the rapid development of artificial intelligence supported by big data and AI technology, in the eyes of researchers, the current development of the robot industry is still in its infancy.

Safety issues: robot technology lags

Qu Daokui, chairman of the China Robotics Industry Association, admitted that in fact, from the perspective of the total amount, the world's robots have developed to the present, and the market is still small. At present, the global average robot density is 0.62%, and China is only 0.3%.

“The robot has been developed for more than half a century. The current global holding capacity is less than 2 million units. In fact, the replacement rate of robots can be neglected. The global average of 99.38% is still homework, and 99.7% of people in China are still working. As an industry, such a small scale is almost negligible," he said.

Many researchers who contacted the reporter believe that the reason is that the robot technology is seriously lagging behind: First, the current industrial robot lacks the perception system, including vision, hearing, etc., the second robot is not smart enough, and a lot of work in the industry depends on Human dexterity, robots do not. Thus, according to human standards, robots are seriously disabled humans and will be excluded from employment. This is the status quo of robotics.

At this level of technology, robots can only do some programmatic and prescriptive work. Work robots that require dexterity can't do it, and work robots with inconsistent work and low standardization can't do it. So in an industry that consumes a lot of labor, robots are almost powerless.

“What is more noteworthy is that robots still have security problems.” Qu Daokui pointed out that robots and humans can't work together now. Robots must work in a cage. This safety limitation puts a lot of work. Excluded.

Turning point from machine to person: focus on perceptual intelligence

Associated with Industry 4.0 is Machine 2.0, the second machine revolution. The typical representative of the second machine revolution is a new generation of robots. This is also the support of the fourth industrial revolution. It must satisfy the connection of objects and objects. Connected to meet the new requirements of data, network, and cloud. This requires a revolutionary change in the robot.

The development of robotic intelligence technology has to go through three stages. The first stage is computational intelligence. At present, robots are mainly at this stage, mainly programming and computer computing. The second stage is to enter the sensing intelligence, through the application of various sensing technologies. Improve the adaptability of the robot to the external variable environment, and enter the cognitive intelligence in the third stage.

Yang Qiang, director of the Department of Computer Science at the Hong Kong University of Science and Technology and president of the Big Data Research Institute and chairman of the 2015 International Artificial Intelligence Conference, told the news that robots have two attributes. The first attribute is mechanical attributes. The robot has so far focused on developing this attribute. The second attribute is the intelligent attribute of the person. At present, the robot industry is in the inflection point from the machine to the person, and the key breakthrough is the perception intelligence to achieve intelligent improvement.

Yang Qiang believes that in March of this year, the battle between Alpha Go and Li Shishi actually brought new enlightenment to the application process of artificial intelligence. Artificial intelligence should have five necessary conditions for success, namely high quality big data. Clear boundary conditions, positive external feedback mechanisms, computing resources, and cross-border data scientists will be the driving force behind the rapid development of artificial intelligence.

"In the next 5 to 10 years, big data-based artificial intelligence, including deep learning and new algorithms extended by deep learning, will receive the first wave of dividends. In addition, small data can also receive deep learning dividends: If you can do migration learning, then there will be artificial intelligence applications in more fields, such as the structure of unstructured data, which is a very broad business prospect; designing a computing architecture specifically for deep learning will change Become a new outlet; software engineering will lower the application threshold of ordinary people; the model of migration learning will learn how to learn, learn learning methods output into a logical expression; natural language is the next big breakthrough direction, that is, can All these unstructured data, natural language descriptions of these pages become expressions in the database, realize the human-machine dialogue system, so in the future we can see a scene where one person leads 10,000 robots to do customer service." .

Researchers generally believe that with the continuous breakthrough of technology, manufacturing is only one of the application fields of the robot industry. The medical, defense security, service and life fields will be a larger space for a new generation of robot applications.

Handheld Pulse Generator

The handheld addresser is used to program the address of the monitoring module offline. When in use, connect the two output wires of the handheld encoder to the communication bus terminal (terminal label 1, 2) of the monitoring module, turn on the black power switch on the right side upwards, and press "ten Add", [Subtract ten", [Add one place" and [Subtract one place" to program the address.

Hd Encoders,Twitch Encoder,Resolver Encoder,Incremental Rotary Encoders

Changchun Guangxing Sensing Technology Co.LTD ,