China State Construction Fifth Engineering Bureau Construction Technology (Xiangtan) Co., Ltd., in collaboration with the Bureau's Engineering Innovation Research Institute's New-Style Transportation Engineering Research Institute, has pioneered the application of 3D + AI intelligent welding robots in the manufacturing and processing of steel structure components.
It is understood that the 3D + AI welding robot consists of a robotic welding system, an electrical vision system, and a robotic walking system. It is equipped with a water cooling system and a torch cleaning and wire cutting system, enabling deep penetration, splash-free, long-duration welding.

Welding robot made in China

3D + AI Welding Robot System Composition

China State Construction Fifth Engineering Bureau Construction Technology (Xiangtan) Co., Ltd. is one of the independent legal entities under the Steel Structure Technology Company of the Third Company of the Fifth Bureau. It is the first intelligent manufacturing base for steel structures designed, constructed, and operated independently by the Fifth Bureau. As an essential part of the Fifth Bureau's prefabricated full industrial chain development strategy, this base has an annual production capacity of over 80,000 tons of various construction heavy and light prefabricated steel components, with an annual output value exceeding 600 million yuan. The base adheres to the product philosophy of "green, intelligent manufacturing, high quality, and research," accelerating the cultivation of new quality productivity with intelligent, digital, automated, and grid-based new technology routes.

The electrical vision system of the 3D + AI welding robot developed this time is equipped with independently developed 3D + AI visual weld trajectory guidance technology. This technology uses innovative MEMS structured light technology, combining low-power MEMS light machines and high-resolution sensors/RGB cameras to achieve fast array scanning, obtaining a complete three-dimensional point cloud (RBGD) of high-fidelity color of the workpiece.

The embedded visual point cloud AI analysis algorithm can identify more than 95% of complex 3D weld features such as lap and butt joints and plan a complete welding process. The automatic welding robot has multi-axis freedom and precise control capabilities, allowing it to perform fillet welding operations in various postures and scenes. Through pre-programming or sensor feedback, the robot can automatically adjust the position and angle of the welding torch to meet different welding needs. The 3D + AI welding robot achieves fully automated high-efficiency welding through rapid virtual 3D modeling of physical objects.

In addition, the 3D + AI welding robot can achieve positioning-free welding in the welding process of steel components. The intelligent welding robot is equipped with a vision system and control algorithm. When steel components are randomly placed on the robot workbench, the terminal component is scanned through the visual surface to perceive the position and posture of the component in real-time. Based on real-time feedback, it automatically adjusts and quickly performs real-time 3D modeling of the entity. Through component 3D model matching technology, it accurately locates the multi-point welding trajectory position of the component without secondary positioning or high-precision positioning tooling. This technology is mainly suitable for factory welding of steel structure components such as lightly loaded structural brackets, stiffener plates for crane girders, stiffening ribs for H-shaped components, and strip plates.

3D + AI Welding Robot Automated Welding

The 3D + AI welding robot also utilizes a non-teaching numerical model weld trajectory guidance welding technology. By constructing a mathematical model of the welding process, it determines information such as the shape, size, and position of the weld. During welding, devices such as laser sensors, vision systems, or torque sensors monitor the position and shape of the weld in real time and feed this data back to the welding equipment. Through control algorithms, the data collected by the sensors are compared and analyzed with the mathematical model, calculating the parameters that the welding equipment needs to adjust, such as the welding torch speed, angle, and welding current. According to the calculation results of the control algorithm, the welding equipment automatically adjusts the position, angle, and movement trajectory of the welding torch to achieve weld tracking and welding. The welding equipment automatically controls the welding parameters to accurately melt and fill the weld.

Welding robot from China

The use of 3D + AI welding robots reduces the time for teaching and debugging, making the welding process more automated and efficient. In the debugging of stiffener plate welding for factory crane girders, taking a 9-meter-long crane girder with 20 stiffener plates on both sides as an example, the total time from 3D modeling to robot completion of welding is about 4 hours. Considering other times such as auxiliary steel component hoisting and flipping by operators, the comprehensive efficiency can reach twice that of pure manual work. One general worker can easily operate the equipment, and the precision scanning camera equipped on the robotic arm can achieve real-time correction of the welding position. This method significantly saves manpower and material resources, reduces welding labor costs, and improves the production efficiency of standard parts such as factory batch production crane girders.

To make 3D + AI welding robot operations a widely used new quality productivity, its cultivation, development, and application will be a long-term task and a systematic project. In the future, Xiangtan Company will adhere to technological innovation as the core, with industrial upgrading as the engine, focusing on the application of 3D + AI welding robots, and striving to find a new growth path with low input of production factors, low resource and environmental costs, and good economic and social benefits, to achieve a revolution in production processes. Through process innovation research, equipment transformation and upgrading, and intelligent platform management, it actively responds to challenges in steel structure processing such as low organizational efficiency, high labor input, inconsistent product quality, and long production cycles, injecting strong "new momentum" into the high-quality development of the steel structure industry.