How does the next robot develop to adapt to the times?

Put on a "high-tech" coat,Robots often appear on the topic of "substituting humans" and make humans beat the drums. As everyone knows, industrial robots that rotate on the production line 24 hours a day are often complained that they are "too stupid", "one bar", "unknown to them".

"Traditional industrial robots can only perform repetitive tasks in a fixed environment and rely on repeated positioning capabilities. They cannot adapt to dynamic and complex environments or cooperate with humans to complete the same work." Zhang Hualiang, associate researcher at the Shenyang Institute of Automation, Chinese Academy of Sciences, told China Science News.

This situation is changing as more and more technologists work to make industrial robots more applicable and easier to use. Recently, a company in the United States has developed a new type of chip. The chip allows the robot to "think" how to handle dynamic scenes, and "imagine" the results of taking different actions, and finally choose the appropriate action to output to the outside, thus giving the industrial robot intelligence.

In the lab, the two robots showed amazing teamwork after they installed the chips. They accompanied each other, passing items from time to time, and trying to get around obstacles that prevented the transmission of items. What impact will the industrial robot move towards intelligence on production and life?

"Simple-minded limbs" is obsolete

Plutonium industrial robots were born in the 1940s, originally for high-risk jobs or for things that humans ca n’t reach, such as nuclear fuel processing. Since then, robots have been used in the automotive, electronics, metal, and logistics industries, engaging in welding, assembly, painting, grinding, cutting, handling, and sorting.

China is a large industrial manufacturing country. Since 2013, China has also become the world's largest consumer of industrial robots. Industrial robots that have been advancing all the way have recently declined, and data from related research institutions show that in the first quarter of this year, the number of multifunctional industrial robots in China dropped by 29% year-on-year. At the same time, data from the National Bureau of Statistics shows that in the first four months of this year, China's total industrial robot output fell by 10% compared to the same period last year.

Zhang Hualiang said that according to the 2018 China Industrial Robot Market Information released by the China Robot Industry Alliance, the decline in demand for electronic and electrical equipment and equipment manufacturing and the automotive industry is the main reason for the decline in the overall industrial robot market. However, the obvious decline in sales is mostly concentrated in foreign-branded industrial robots. The sales volume and share of China's own-brand industrial robots have actually increased compared to previous years.

In the face of the decline in total output, he believes that the industry needs to pay attention to the adjustment and upgrade of the research and development of industrial robots. "For example, complex programming has greatly limited the scope of use of industrial robots, leading to a narrower application of industrial robots."

How to find a breakthrough in the development of industrial robots? Zhang Hualiang said that the application fields and industries of industrial robots should be widened, and the application potential of SMEs should be actively explored.

Meikamand company founder and CEO Shao Tianlan is more simple and clear, industrial robots are the key basic technology to achieve intelligent manufacturing. For industrial development, it is the only way to give industrial robots intelligence and increase the application scenarios of industrial robots.

"Checking the color" is not enough

What is intelligence? Shao Tianlan gave an example. If the air conditioner in the home can be automatically adjusted according to the temperature and number of people in the room, when it finds that there are elderly and children in the house, it can automatically reduce the wind, and when it finds a few sweaty young men in the room, it increases the wind. This shows that the air conditioner has intelligence.

"The important feature of intelligence is to perceive the environment and adjust it according to the environmental conditions. Therefore, the important perception ability to realize the primary intelligence of industrial robots is vision, followed by hearing, and other senses add up less weight than these two."

He believes that to achieve the vision and hearing goals of the robot, there are two challenges in terms of hardware: one is the sensor, especially the visual sensor; the other is the algorithm with intelligent decision-making.

At present, research represented by perceptions such as vision and hearing has become a hotspot in the development of industrial robots. But Miao Li, founder and CEO of Cooper Technology Co., Ltd., doesn't think so. "Although the development of vision, force control, and tactile sensors is the core business of Cooper, the intelligentization of industrial robots is a system-level problem. Industrial robots are a multi-disciplinary cross-integrated product. On the other hand, you can't just focus on sight or touch. "

To pay attention to the design of industrial robots, we must first pay attention to the original design-this determines the scope of the robot's capabilities. Secondly, the robot needs to have the ability to make decisions. For example, when a ping-pong ball flies over, the robot needs to determine whether the ball comes up or down, whether it is a forward shot or a back shot, where the attack origination point and the landing point are. After perceiving, it is necessary to judge the state after it bounces. After a series of "thinking", the robot executes the result of thinking and hits the ball in the future.

Do you think it's over? In fact, there is one final key step-introspection. Li Miao said that the robot does not always hit the ball beautifully, and there are also "misses", which requires it to find problems in introspection, learn lessons, and perform intelligent automatic upgrades.

"Smart" love industry

At present, intelligent industrial robots are still in the laboratory research stage, "striving hard" on the basis of traditional robots to improve application performance.

In the first-line years, Li Miao's feelings are more specific. He believes that most of the research on industrial robots stays at the stage of design, perception, and precise execution. There is less research on robot decision-making judgments. "As a closed-loop research team is even rarer.

Although it is still in the conception and research stage, the R & D personnel have a glimpse, which is enough to imagine the future application scenarios of industrial robots.

Shao Tianlan said that intelligent industrial robots will widely appear in industries with low automation, such as logistics, and human-intensive industries, such as clothing, shoes, automobile assembly, and toy assembly.

According to Li Miao, they are developing a robot for picking mushrooms. Although the shiitake mushrooms grow in the same place, their patterns are different and they need to be classified. In the past, manual sorting of shiitake mushrooms gave the industrial robot the "sorting" vision, and the robot could judge the shiitake mushrooms.

Industrial robots are given more intelligence, and its application range will be wider. "In the future, there will be more and more application scenarios for industrial robots. The technology of industrial robots will also be used in service robots, assistive robots, elderly robots, etc. that are needed in daily life. Industry. "Zhang Hualiang said.

Free human labor

Talking about whether industrial robots will replace human work, Zhang Hualiang said that it is certain at some levels. For example, in a factory assembly line, a sorter who selects products by category, a worker who grinds components, and a quality inspector who checks the quality of the product may be replaced by industrial robots. "From a positive perspective, as humans are replaced by industrial robots, they can urge humans to find new employment directions. Encouraging humans to increase their work skills and possess more complex work capabilities, this transformation is meaningful for human development."

Li Miao said that although robots will replace some people, most of these tasks are repetitive, heavy, and even toxic and harmful physical labor. Instead of "replacement", it is better to say "liberate" and liberate manpower. When low-end jobs are replaced by robots, labor costs will decrease, and better jobs will emerge in China.

He emphasized that in addition to paying attention to whether it would "replace humans", we should also pay attention to the changes that industrial robots have brought to industrial data. The realization of intelligence must pass data, and the core of intelligence is data-based decision-making. The process of industrial robot intelligence is also a process of making industrial data more valuable. This is the great contribution of industrial robot intelligence.