As the most commonly used connection method during assembly, screw lock is a high-precision and repeatable operation, and is widely used in all aspects of processing and manufacturing. When the traditional manual screw lock operation is performed, problems such as loose screws and unstable locking are prone to occur, which affects the safety and life of mechanical equipment. At the same time, the cost of manpower production is high and the production efficiency is low, and 24-hour continuous production cannot be achieved.
Therefore, using machine vision equipment instead of manual screw locking is the best choice for manufacturing enterprises. The use of machine vision equipment for screw locking not only liberates heavy manual labor and avoids the occurrence of product quality inconsistencies caused by human factors, but also improves product production efficiency and processing from traditional manual tightening to the latest visual positioning technology. The quality has been improved in both directions, so that machine vision positioning technology can enter the field of automated manufacturing on a large scale.
3D vision technology to realize vision-guided screw locking
Machine vision means that the machine replaces the human eye for measurement and judgment. In the manufacturing process, parts often have defects, and quality inspection of related products is required.
The machine vision system analyzes and digitizes the captured image, and then judges the size, shape, color, etc. based on information such as pixels, brightness, and color. According to the results of the judgment, it can not only detect holes, scratches, dirt, and damage It has high accuracy and work efficiency, and can meet the needs of mass production inspection.
With deep learning + traditional machine vision technology as the core, Shenmo Technology innovates and develops industrial AI vision system. Through the visual positioning and defect detection functions of the machine vision system, it is equipped with 3D vision technology and a variety of application algorithms to deal with various locks. Scenarios, thereby improving the flexibility and flexibility of processing, and realizing the all-round empowerment of vision-guided screw locks.
Vision-guided screw lock payment case
Background of the project
The locking product is the mask of the air-conditioning outdoor unit. The front-end process is manually completed by manually fitting the mask and the cover plate to the screw holes and then locking the screws. Various defects are difficult to be detected by traditional visual operators.
Project Difficulties
1. Various defects such as mask falling off, misalignment, and hole blocking
2. There are large differences in hole position characteristics
3. It is necessary to correct the deviation before locking and paying
Technical advantages
1. Using multi-scale feature fusion learning algorithm and confrontation reconstruction model
2. Strong anti-interference ability, high stability and wide compatibility
3. Adopt hole position evaluation strategy and hand-eye coordination dynamic correction
Project effect
1. It can efficiently calculate the deviation posture of the multi-layer hole position and the best screw locking point
2. In the case of hole misalignment, dynamically adjust the position and posture of the screwdriver to correct the hole position and lock the point
3. It can meet the detection of the installation status of screw holes in various situations
Industrial AI vision system to meet various requirements of users
1. Meet the user's requirements for high precision
The requirements for screw locks are strict, and the comprehensive positioning accuracy is high, which puts forward higher precision requirements for machine vision equipment. Based on 2000+AI industry models, the industrial AI vision system can quickly complete algorithm model matching, data labeling, training, and production line version updates. While satisfying the iteration of defect types in the production process, the defect detection accuracy reaches the pixel level, meeting the stringent requirements. Accuracy requirements.
2. Meet the user's requirements for stability
The use of machine vision systems to replace manual screw locking operations can avoid manual errors due to fatigue, inexperience and other unstable factors during processing. The industrial AI vision system can effectively reduce the interference of external environment changes on the detection, achieve the consistency of the assembly process, and realize precise scanning through the 3D camera, and guide the robot to accurately implement the tightening process.
3. Meet the user's requirements for high efficiency
The number of screws on most processed parts is large, and the efficiency of manual tightening is limited. The industrial AI vision system, through 3D vision to guide the robot to perform repetitive locking operations, can realize 24-hour continuous operation according to the production plan, and the efficiency has been significantly improved. At the same time, due to the higher degree of standardization of mechanical operations, missing or wrong screwing rarely occurs, rework will be greatly reduced, the speed of product off-line will be accelerated, and the processing efficiency of the production line will be further improved. Time cost and material consumption cost will also be reduced. And then down.
Facing the needs of the industry, the industrial AI vision system developed by Shenmo Technology based on 3D vision has realized the precise operation and defect detection of screw lock operations, enabling related manufacturing companies to quickly enhance the production line at a lower time cost The automation, intelligence and flexibility of processing and assembly can improve production efficiency and achieve the purpose of reducing costs and increasing efficiency.