Ursa major intelligent technology

Ursa major intelligent technology | Does intelligent welding really have hands?

It is a long-term development goal of science and technology to liberate human beings from heavy labor with automated machines. With the increasing cost of welding labor and the increasing gap of skilled talents, enterprises have a strong demand for robots to fill the talent gap. According to the data of Paisi and Guangfa Securities Development Research Center, the sales volume of domestic welding robots will exceed 50,000 units in 2020. However, the welding robots on the market are basically in an automatic state and are not intelligent, so they are not suitable for many high-end manufacturing fields.

Does constellation intelligent technology intelligent welding really have hands? " img _ height = " 68 1 " img _ width = " 960 " data-src = "//img q8 . q578 . com/ef/07 14/6a 3c b 62260982 DCD . jpg " src = "/a 2020/img/data-img . jpg " >

The existing welding robots in the market generally need manual teaching to determine the welding trajectory. Because each action needs manual teaching once, it is extremely time-consuming. In addition, it is extremely difficult to teach and program for high or deep welds that are difficult to reach by manpower. In addition, the programming languages of mainstream brands of robots are different, and it takes at least a few weeks to master the operation of a robot. At the same time, the process parameters can't be adjusted, the motion trajectory can't be modified, and the welding program can't be adjusted, which leads to the need for the workpiece to adapt to the working conditions of the robot in actual use, which is extremely demanding and difficult to meet, and the actual welding efficiency is less than 30% of the theoretical value. The fault tolerance rate of automatic teaching programming for workpiece machining accuracy is extremely low, and it is difficult to meet the needs of most other industries except the automobile industry with high standardization. The market urgently needs intelligent welding robots that can meet the needs of manufacturing industry.

Does constellation intelligent technology intelligent welding really have hands? " img _ height = " 622 " img _ width = " 1080 " data-src = "//imgq 8 . q578 . com/ef/07 14/6a 5a 30 f 6896 18 Fe 1 . jpg "src ="/a2020/img/data-img.jpg "> Welding is a chaotic system with complicated process and changeable situation. From the application perspective of AI technology, it can be compared with the familiar autopilot. Both of them need the ability to execute according to the preset program, and also need the real-time perception and analysis ability to deal with frequent emergencies. The difference is that the decision of autonomous driving is only reflected in several processing forms such as braking, acceleration and lane change, so in the intelligent form, autonomous driving belongs to "emphasizing perception and ignoring decision-making".

Let's take a look at AlphaGo again, which is easier to understand from another angle. Based on Go 1.43* 10768, many results of the game need great data processing ability and decision-making ability; Compared with the powerful decision-making ability, AlphaGo's perception ability is relatively weak, and it only needs to scan to determine the actual position of the landing point, so AlphaGo belongs to "decision-making over perception".

In actual welding work, intelligent welding sensing system needs to realize digital sensing modeling of workpiece and molten pool under harsh conditions. In order to ensure the high accuracy of welding results, small changes will cause huge data changes and processing, which is finally reflected in the intelligent welding robot's perfect completion of multi-layer multi-pass planning, variable gap welding, staggered welding, thin plate welding and other complex and difficult welding work, which is what we call whole-process adaptive welding, which is why only intelligent welding robots can replace highly skilled welders in the end.

Does constellation intelligent technology intelligent welding really have hands? " img _ height = " 805 " img _ width = " 1080 " data-src = "//img q8 . q578 . com/ef/07 14/e 0 a 2 efb 2 ecae 872 c . jpg " src = "/a 2020/img/data-img . jpg " >

It can be seen that it is quite difficult to call it intelligent welding, and it is not necessary to have a "hand". In order to realize adaptive welding, it is necessary to design a system according to the welding process requirements, drive the system to adapt to the situation of the workpiece, replace the work logic that the workpiece needs to adapt to the robot in the past, establish the identification model of various workpieces and the process database of various materials, and realize the dynamic parametric control of the welding process. At the same time, the sensing system needs to realize the digital sensing modeling of workpiece and molten pool under harsh working conditions, and the multi-sensor fusion scheme is essential.

Does constellation intelligent technology intelligent welding really have hands? " img _ height = " 506 " img _ width = " 1080 " data-src = "//img q8 . q578 . com/ef/07 14/35 a2 BC 30 da 0 ef 7 c 9 . jpg " src = "/ A 2020/img/data-img. jpg "> The welding difficulties faced by high-end manufacturing in marine engineering, bridge construction, aerospace and other fields are usually reflected in two aspects: the small batch and multi-specification shape of workpieces and the complexity of welding. These two factors combine to form the welding problem of manufacturing enterprises in various fields.

Does constellation intelligent technology intelligent welding really have hands? " img _ height = " 656 " img _ width = " 786 " data-src = "//imgq 8 . q578 . com/ef/07 14/40 e 9 c 7 CD 1 19e 5635 . jpg " src = "/a 2020/img/data-img . jpg " >

In the bridge steel structure prefabrication industry, the diversity of workpieces forces the users of traditional welding robots to face robot programming problems for several days before each workpiece switch, which leads to the need for enterprises to recruit professional robot engineers to complete complex robot programming and cannot fundamentally solve the automation application problems. In the application of China Railway Yangzhou Baoqiao steel structure production base, the welding master of Ursa major realized self-perception to judge the shape of the workpiece and generate the required robot trajectory. The trajectory generation time is shortened from 5 days to 30 seconds, which greatly improves the production efficiency of the corresponding process and solves the automation application problem caused by structural diversification in heavy industries such as steel structure for a long time.

Does constellation intelligent technology intelligent welding really have hands? "img _ height =" 406 "img _ width =" 725 "data-src ="//imgq8.q578.com/ef/0714/c817E2f08f3a3782.jpg "src ="/.

In the difficult non-ferrous metal pipeline welding project of China Ocean Shipping Group, the biggest problem is to solve the welding problems caused by raw materials and assembly errors. In this project, the welding master of Bear analyzed and judged the workpiece at a speed of 200ms, and called various process parameters in real time, thus realizing the real whole process adaptive welding. The qualified rate after welding is 98%, which is 4 times that of manual 1.4 times. At the same time, it can produce 120 inch pipe section every day, and the production efficiency is five times that of workers.

Does constellation intelligent technology intelligent welding really have hands? " img _ height = " 605 " img _ width = " 1080 " data-src = "//imgq 8 . q578 . com/ef/07 14/e 5 1 c 44 ea 228 b 1577 . jpg " src = "/a 2020/img/data-img . jpg " >； WELDING MASTER, an intelligent welding robot product introduced by Daxiong Intelligent Technology Co., Ltd. (hereinafter referred to as Daxiong), has the characteristics of independent programming without manual teaching, high fault tolerance rate of workpieces, wide application range of various materials, high welding efficiency and good quality, and is suitable for small batch and diversified workpiece production. Its welding technology can reach 10 years welding experience, which can solve the problem of shortage of welding skills talents for manufacturing enterprises.

Does constellation intelligent technology intelligent welding really have hands? " img _ height = " 608 " img _ width = " 1080 " data-src = "//imgq 8 . q578 . com/ef/07 14/ 1 e 573 fc 64 c 8535 1f . jpg " src = "/a 2020/img/data-img . jpg " >；

As the welding robot with the highest degree of digitalization in the industry, the welding master is equipped with a new generation of WOS intelligent welding system independently developed by Ursa Ursa, which completely breaks through the technical barriers and realizes the full-specification and full-process adaptation of welding workpieces. The self-programming time is as fast as 30s, and the fine process control of 20ms and 0. 1mm is realized. Software-driven welding masters can deal with special situations that occur randomly in the welding process like experienced welders, and the welding qualification rate is as high as 98%. At the same time, it is suitable for the change of complex process parameters in the welding process of special materials in high-end manufacturing field, and the qualified rate of welding products is much higher than that of general welding robots.

At present, the welding master products have been applied to FPSO shipbuilding in addition to cooperation with COSCO Heavy Industry in shipbuilding field. At the same time, they have won the bid for CIMC Pacific LNG spherical tank and stainless steel pipe section welding project of China Merchants Heavy Industry, and conducted case verification tests with China Aerospace and China Petroleum. Ursa major hopes to achieve the role of benchmarking by cooperating with leading enterprises in the industry. In the future, Ursa Major plans to expand into aerospace, high-end equipment, energy equipment, rail transit and other fields, and it is expected to be applied to about 10 industries.

Does constellation intelligent technology intelligent welding really have hands? " img _ height = " 606 " img _ width = " 1080 " data-src = "//img q8 . q578 . com/ef/07 14/2977442 ef 5b 2 f 56 a . jpg " src = "/ A 2020/img/data-img. jpg "> The bear constellation was founded on February 2019,65438. With the mission of "improving the working environment of 20 million welding workers", the company actively promotes the digitalization and modularization of welding master experience. The intelligent welding operating system independently developed by the company integrates environmental awareness, target analysis and behavior decision-making, and has been adapted to multi-brand and multi-structure industrial robots, and has been applied in many fields such as ship and ocean engineering, aerospace equipment, bridge prefabrication, rail transit and so on.

Does constellation intelligent technology intelligent welding really have hands? " img _ height = " 1 19 " img _ width = " 1080 " data-src = "//imgq 8 . q 578 . com/ef/07 14/af 89 a 83 f 69 ab 768 1 . jpg " src = "/a 2020/img/data-img . jpg " >