Because with the improvement of its processing capacity, the cost of processor and memory continues to decrease, programmers can do more work, not only shoot an image, but also compare it with the corresponding reference model, and most of the early 2D visual images are based on this technology.





Development of vision system



Robot vision system was first applied to the complete body production robot of automobile production line. If there is no vision system, special tools need to be used to process the aperture on the car body, so that the robot can know the specific position of the car body.





However, after the development of visual camera system, such expensive processing tools are no longer needed.



The robot can automatically determine the exact position of the vehicle body, and then calculate the positions of the four apertures mathematically.





Mr. Roney of FANUC further explained: "from a very early time, we realized that the use of vision system is a good way to reduce the cost of tools. Processing tools are very expensive and fixed. If the manufacturer needs to change the model in the second year, all tools must be updated." According to Mr Roney, after adopting vision technology, automobile manufacturers can save millions of dollars a year in mold and fixture manufacturing.





Working principle of vision system



The vision system uses computing technology to recognize images, and then teaches robots to recognize things and find what they need through training. The image is based on a large number of pixel data. Each pixel in the series has a gray level, and then the data is analyzed by operation technology.





The robot can judge the shooting position of the image, so it can recognize the position of the object, and then judge its corresponding size, shape and quality; The program can also be changed according to the image and algorithm. For example, parts of different sizes can take different paths, and part a will fall where part B should fall.





David dechow, President of aptura machine vision in Lansing, Mich., explained: "robots with vision systems will eventually be able to manipulate any type of parts in any direction.





”Using the attached camera system or remote control camera, we can obtain the fast captured image of an object and find the relative position between the object and the robot.



"Then, the FANUC robot can use this position data to determine the position of the object, no matter where the robot is moved," said Jeremy Pennington, a control engineer at guideengineering in Ft. Wayne, Ind





Application of 3D vision system



3D vision system can be applied in many fields, among which ideal results have been achieved in the fields of picking parts in containers, workpiece loading and unloading on machine tools, packaging and welding.





1. Used for picking parts in containers



So far, robots are mainly used to pick up parts stacked arbitrarily in containers. To achieve this goal, we need to meet three basic elements: vision system, container anti-collision and anti-collision monitoring system. There is no doubt that the visual system is needed, because we first need to find out where the parts are. However, only relying on the container wall for detection has certain limitations, because the robot claw will further penetrate into the bottom of the container, so the picking of parts will be more and more difficult. The modeling method used to pick up parts in the container can fully understand the limitations of picking tools, sensors and manipulator. Under such working conditions, once the position of the part is determined, the robot starts automatic calculation to determine whether it can really pick the part from the container. The third element is anti-collision detection. Finally, the manipulator will inevitably touch the container wall, so it needs to distinguish whether it belongs to soft contact or hard contact. If it is hard contact, it may damage the robot system.





Mr. Roney of FANUC said: "we have achieved great success in using robots to pick parts structurally and arbitrarily in containers." The so-called structural picking parts means that each part in the container is facing upward, while arbitrary picking parts means that the parts are stacked arbitrarily in the container. Mr. Roney believes that the latter part picking method is more challenging, but if the above three elements (vision system, container anti-collision and anti-collision detection system) are available, everything can be achieved.





2. It is applied to workpiece loading and unloading on machine tool



"In many applications, after picking up parts, they are directly installed on the machine tool for processing. In most cases, the fixture system of the machine tool does not allow any errors in the installation position of the robot's parts," said Mr. Burg of Ellison technologies Therefore, the accurate positioning of parts is very important for the clamping system of parts.



The robot can determine the position of image shooting, so it can recognize the position of the object, and then make corresponding judgments on the size, type and quality of the object.





Without these functions, the machine tool may suffer heavy losses. Therefore, the use of vision technology is conducive to the correct positioning of parts.





3. Application in packaging field



Vision technology is also very key to the field of packaging application. For example, food often needs to be transported to the food picking area through conveyor or slider device, so there is no problem of repeated positioning. But in fact, because the food is in different positions, there needs to be a process of reorientation and picking, and then put it into the packing box. The vision system can help the robot find the corresponding food and complete the packaging task according to the requirements.





4. Application in welding field



When welding, the robot can use the vision system to adapt to the subtle changes between the two welding elements. Even in the field of spot welding applications, visual technology can be used to correct errors.



Application of the system in practice



Ellison technologies automation has installed a robot system with 3D vision technology for processing plates. The plate with a long service life is welded into a barrel by the robot, and then the material is transported into a barrel with a long service life.





"The welding process may contaminate the clamp. If the clamp touches a new part, the part will be contaminated," Mr. Burg explained Before using visual technology, we have to adjust it manually. Now, each position on each part can be verified. "This technology allows the robot to display the condition of the whole part through a 3D camera system," Mr. Burg continued. "It allows operation without human intervention."





In the 10 years before the adoption of vision technology, the company had four FANUC robot systems, and at least one robot system often needed manual adjustment. Now, the company only needs three sets of robot systems to complete the same workload, and there is little need for manual intervention.





The future of 3D vision system



What will happen to the future robot 3D vision system when the production cost is further reduced and a more powerful system is adopted?



Mr. Roney of FANUC replied: "it may enter the so-called visual servo system that is more interesting. So far, the visual system we are considering is only looking for an object in time through a certain point." However, the parts are often in motion. At this time, the effect of using single point on-time and fast camera may be poor; The visual servo system can continue to obtain a variety of information about the position of the object. This system allows the robot to guide and know the position of the object at all times.





"Sooner or later, this system will be applied in the assembly line, because in the assembly line, the assembly parts are always suspended on a running transmission chain or other material conveying devices," Mr. Roney suggested Even without further development, it is obvious that the processing capacity of this visual servo system will continue to improve, and the production cost will gradually decline. In the next few years, using this combination can make 3D vision technology more widely used.