Increasingly, manufacturing is moving towards Industry 4.0, where factories rely more on automated processes instead of on traditional human labor. Against this backdrop, more and more manufacturing plants are employing machine vision, which can be thought of as the “eye” of production.
Increasingly, manufacturing is moving towards Industry 4.0, where factories rely more on automated processes instead of on traditional human labor. Against this backdrop, more and more manufacturing plants are employing machine vision, which can be thought of as the “eye” of production and can come in handy in various operations, for example inspection or guiding automated guided vehicles (AGVs).
Basically, machine vision addresses the fundamental issue that the human eye has limits. As defects on products, for example textiles, can be infinitesimal, they can be hard to detect by the human eye, especially when it gets tired or fatigued. Moreover, humans see things differently; 10 people working at a production line may each apply a different standard to inspecting products, rendering inspection ineffective.
This is where machine vision can help. “Machine vision systems will continue to be the primary consideration for manufacturers who are looking to improve quality or automate production. They are built for visual inspection and control under demanding industrial applications that require high-speed, high-magnification, 24-hour operation, and/or repeatable measurements,” said Bruno Menard, Program Manager for Embedded Vision at Teledyne DALSA. “Machine vision has the ability to perform repetitive tasks faster and more accurately, with greater consistency over time than humans. In addition, machine vision can reduce labor costs, increase production yields, and eliminate costly errors associated with incomplete or incorrect assembly.”
Machine vision can be employed in different manufacturing operations, including inspection, orientation, and robotics or AVG guidance.
Inspection remains the most important machine vision application in factories, whereby the vision system can check for broken stitches, stains or other defects that are too tiny to be detected by the human eye. “The biggest influence on smart factory from machine vision technology is inspection area. Machine vision provides efficient inspection through image analysis of the products obtained in the process,” said Hongsuk Lee, Business Development Manager at SuaLab. “Depending on the industry and products, some have built a complete automation of the inspection process through machine vision, and some have used the machine vision technology to significantly reduce the number of manpower in the inspection process if the inspection system is not fully automated.”
Another operation that can benefit from machine vision is orientation. “With pattern-matching, we can determine position, orientation and scale of a part or object. After training on a ‘golden pattern,’ algorithms are able to move, rotate and scale (up or down), in preparation for taking a measurements always at the same place. After setting a threshold for acceptance, the application is able to find the object, bring it to the original position and compare it, and finally assigns a pass or fail,” Menard said.
With robotic arms and AGVs increasingly deployed in factories, machine vision can also help. For example, when a robotic arm picks up something from point A to point B, a depth camera can provide the necessary “vision,” telling the robotic system where it is to pick up the particular object and where it is to place it. For AGVs, the operator no long needs to set up tracks for the vehicles; a depth camera integrated with the AGV can tell the vehicle where it is now, where it is supposed to go, and what obstacles to avoid along the way.
Manufacturers that can benefit from machine vision aren’t so vertical specific – most industries from traditional to electronic can leverage what machine vision has to offer. “Machine vision is a primary consideration for any manufacturer who is looking to improve quality or automate production. Multiple industries such as semiconductor, flat panel display, electronics, automotive, food and packaging and general manufacturing can benefit from implementing a machine vision system to improve productivity and enhance customer satisfaction through the consistent delivery of quality products,” Menard said.
“From complete products to assemble parts, the last step in every manufacturing process is the inspection process, so technically all industries can benefit from machine vision technology,” Lee said. “However, since there is a difficulty in image acquisition process for each product, the existing machine vision technology has been applied to specific industries like electronic industries (semiconductors and displays) which are easy to obtain images. If we overcome these limitations with deep learning technology, we will be able to benefit from machine vision technology in any manufacturing industries.”