Flexible Intelligence: a watershed in the evolution of industrial robots
2022-01-19
Robot track is hot in 2021. According to incomplete statistics, in November alone, there were more than 17 related financing projects in the robot field, and the financing projects of more than 100 million yuan exceeded double digits. Under the crazy pursuit of capital, the robot industry ushered in a new round of outbreak in 2021. A large number of front-line investment institutions intensively scanned the robot track, and new projects have sprung up. In a longer time line, China's robot industry has developed rapidly, with a compound annual growth rate of about 15%. In 2020, the operating revenue of the robot industry has exceeded 100 billion yuan. Among them, the output of industrial robots has exceeded 210000 units. According to the data of the Ministry of industry and information technology, the density of manufacturing robots has reached 246 units / 10000 people, twice the global average. In the industrial field, the process of replacing people with machines began to change from quantitative change to qualitative change. According to Yang Yiming, founder and CEO of Moying technology, the large market of robots in the future will appear in deep application scenarios that need more high-end and intelligence. The evolution of industrial robots is at a watershed "Since around 2013, China's robot industry has developed very rapidly, and entrepreneurs and scholars have constantly put forward the view that" one year is the first year of robot ". For example, many people regard 2013 as the first year of robot development, because China is the world's largest industrial robot market since that year; others regard 2021 as the first year of robot, because the epidemic has put everyone's attention to the machine The cognition of human beings has been pushed to a new level. " Yang Yiming told China Electronics News, "at present, technology and industry are developing. Industrial manipulator, cooperative robot, AGV, AMR, meal delivery robot, explosion-proof robot and other products are increasing and the scale is expanding. Correspondingly, the in-depth fields of robot are also increasing, including 3C, semiconductor, biomedicine, aerospace and other fields." However, compared with the foreign advanced level, China's robot industry still has some problems, such as insufficient technology accumulation, weak industrial foundation, lack of high-end supply and so on. The recently released "14th five year plan" for the development of robot industry proposes that by 2025, China will become the source of global robot technology innovation, high-end manufacturing agglomeration and a new highland of integrated application. A number of robot core technologies and high-end products have made breakthroughs, the comprehensive indicators of the whole machine have reached the international advanced level, and the performance and reliability of key parts have reached the international level of similar products. Since the birth of remote control manipulator and master-slave manipulator, robot has developed for more than 70 years. Analyzing the historical evolution and future development of the robot industry, experts believe that from the perspective of robot capability, it can be divided into six stages from l0 to L5, and l0 to L2 covers the state of all existing robots. L3 level robots will become a watershed for the improvement of robot intelligence and lay a foundation for the realization of L4 and L5 level robots in the future. Yang Yiming said that we take history as the main line and the function of satisfaction as an important angle to infer new trends. In 1940s, in order to deal with special scenes (such as nuclear radiation), remote control manipulators and master-slave manipulators were born one after another. The intelligent level of robots was very low, and people needed remote control operation. At present, medical surgery robots and UAVs still belong to this category. Around 1954, the concept of industrial robot was put forward. The intelligent level of robot was relatively improved. The robot can realize repetitive single fixed action based on mechanical and simple electronic signals. This is also one of the most widely used robot categories. Around 1966, mobile robots with sensing ability appeared. They can use sensors to measure environmental information and complete a single action. In recent years, mobile robots have been widely used in industry, e-commerce and other fields. L2 level machines have been put on the market, including on-site logistics, unmanned intelligent warehousing, etc. The characteristic of this kind of scene is that when the environment changes slightly, the robot can complete the established and single action through the corresponding sensor signal or prefabricated algorithm. Since 2013, L3 level flexible intelligent robot has been developed. Yang Yiming believes that 2022 will be the first year of mobile cooperative robot, and L3 flexible intelligent robot will be available. It also has the ability of perception. It can use various sensors to measure environmental information, identify, understand and feed back preset actions through preset programs. Compared with traditional robots, L3 level robots are not only the evolution of existing robots - doing simple and repetitive work instead of humans, but also an upgrade to stronger intelligence. They change from serving "humans can't do and don't want to do" to serving "humans can't do". For example, flowline and CNC (data machine tool) loading and unloading are all use scenarios. In the future, at higher L4 ~ L5 levels, the focus of robots will further tend to serve "humans can't do it". Robots will also evolve from simple "machines" to "intelligent robots", which are more intelligent and have human attributes. Robots will be more flexible and intelligent According to this evolution direction, what will the future robot be like? Yang Yiming believes that robots in the future will be more flexible and intelligent. In the past, most new applications of industrial robots needed code. After the deployment of the same type of scenario, it is often unable to be directly reused in other similar products. L3 intelligent robot can quickly realize secondary development and reduce marginal cost. Yang Yiming said that through innovative design, Moying technology has realized hourly secondary development. Users realize zero code second deployment through ink shadow Fuxi teaching pendant and drag and drop engineering module. On the basis of no need to knock the code, you only need to drag and drop the module on the teaching pendant to complete the deployment and adjustment of complex scenes. Ordinary factory workers can also operate through simple training. The simple deployment of L3 robot fundamentally solves the marginal cost requirements of flexible manufacturing. In the future, there is also a trend for robots to be super multi machine and super flexible. Taking numerical control machine tools (CNC) as an example, in the large-scale CNC scene, there are nearly 10000 CNC equipment in the plant, which are operated by hundreds or even thousands of ordinary workers. If you want to realize that robots can replace ordinary workers, you must realize super multi machine and super flexible cooperation. Hundreds of robots and more than 1000 robots need to be the same as people. They can divide labor and cooperate to realize different processes and operations. "For example, if you want robots to help you dress, two ink shadow mobile cooperative robots work together. One robot is responsible for unfolding the coat, and the other robot raises the corners of the clothes from the ground to dress you in an adaptive way through the cloud scheduling system. The cooperation ability is comparable to that of human beings." Yang Yiming said, "with one robot looking after 10 devices, in the scene of 10000 machine tools, 1000 ink shadow robots work together in the factory like thousands of general workers." From the demand of the manufacturing industry, because the products produced are becoming more and more complex, robots in the future are bound to be able to move and operate autonomously like people. In short, they can move and operate like people. Therefore, under the requirements of mobile cooperation, the number of joints and degrees of freedom of robot ontology will increase greatly, which also means that the data flux and algorithm complexity of robot brain will face great challenges. The intelligence level of robots will also be higher. In the past, robots bought can only be depreciated with their service life. With the use of OTA (over the air download) technology, L3 intelligent robots can realize their continuous evolution and growth through OTA technology under the future oriented ultra complex OS system technical architecture. Mobile cooperative robots used in CNC scenes can use OTA technology. Buyers can import new scene modules to realize other scene applications, such as ice cream making, coffee brewing, etc. a robot can have "18 kinds of martial arts". At the same time, it also enables the robot to learn and grow by itself. Flexible intelligent robot has better cost performance "Column 3" of the "14th five year plan" for the development of robot industry defines the key development direction of high-quality industrial robots, including high-precision and reliable welding robots for automotive, aerospace, rail transit and other fields, and vacuum (clean) robots for automatic handling, intelligent movement and storage in the semiconductor industry, Civil explosive production robots with explosion-proof function, AGV, unmanned forklift, sorting, packaging and other logistics robots, large load, light, flexible, double arm, mobile and other cooperative robots for 3C, auto parts and other fields, which can move at any position in the working area such as transportation, grinding and assembly, and realize the reachability of any position and attitude in the space Mobile operating robot with flexible grasping and operation ability. Yang Yiming told China Electronics News that as the founder of L3 intelligent robot "mobile cooperative robot", inkshadow technology focuses on the loading and unloading of CNC generation processing of 3C / automotive electronics. The reason why we focus on this market is that there is a large market demand on the one hand. Public data show that at present, the gap of operating workers in the CNC industry is as high as more than 600000. Based on the average monthly salary of 7000 yuan, the annual comprehensive employment cost of a general worker is about 150000 yuan. Although the cost is not low, young people are not willing to engage in this kind of work, and the job mobility is very large. On the other hand, this scenario needs innovation and high-end and intelligent devices. It is roughly estimated that the sales volume of domestic CNC machine tools will be more than 50000 in 2020, and the total volume will be close to 100000 with brands outside Shanghai. According to the ratio of one ink shadow mobile cooperative robot to 10, the annual demand for robots is as high as more than 10000. However, intelligent and high-end robot manufacturers in this market do not exceed double digits. Yang Yiming said that the future must be the era of software. At present, Moying has developed a set of modular and expandable intelligent robot operating system, which can maximize the intelligent level of the robot. The operation of the robot has changed from inputting operation instructions through programming steps to only issuing tasks, and the robot completes all steps at one time, which is more flexible and efficient. The system was initially applied to mobile cooperative robots, and the final idea is that it can be applied to all types of robots in various industries. "Just like the relationship between mobile phones and Android and IOS operating systems, this operating system can be used in robots produced by major manufacturers." Yang Yiming said, "this system is currently used in ink shadow's own products, but it may be opened to the outside world at a certain node in the future." (Xinhua News Agency)
Edit:Li Ling Responsible editor:Chen Jie
Source:CENA
Special statement: if the pictures and texts reproduced or quoted on this site infringe your legitimate rights and interests, please contact this site, and this site will correct and delete them in time. For copyright issues and website cooperation, please contact through outlook new era email:lwxsd@liaowanghn.com