Machine Vision: An Overview
Machine vision has become an integral component of automated manufacturing processes. In older automated manufacturing lines, less efficient methods were used to corral a product part down a conveyor and into a precise position for fixation. This more manual process often required additional hardware to corral pieces down the assembly line in the proper direction or would require pre-fixturing a part with manual labor. This not only was more time intensive but also made manufacturing processes more susceptible to alignment errors.
Machine vision uses cameras to locate product parts throughout the entire manufacturing process. Once the camera has identified a part, it sends the part’s positioning information to the machine. This enables the machine to pick the part up and secure it for the next step in the manufacturing process.
Machine vision is especially useful for identifying and tracking parts with complex shapes. Machine vision is also used in tandem with barcodes, QR codes, and other traceability methods. It's also useful for inspecting parts before fixturing them throughout the manufacturing process—limiting the number of faulty parts making their way into the final product.
Machine vision systems employ either 2D or 3D cameras to conduct live-time iteration, part inspection, and increase efficiency. Both offer unique benefits, based on the applications they’re used in. In this article, we’ll look at the key benefits of using machine vision in automation as well as the scenarios where 2D and 3D vision can be best utilized.
Benefits of Using Machine Vision in Automation
On the whole, machine vision is much more efficient than manual sorting and fixturing processes. Machine vision is also more accurate, giving the larger machine a better sense of part alignment and type.
Machine vision technology allows for more flexibility in the manufacturing process as well. It mitigates the need for new hardware or the development of new manual processes any time manufacturers want to use an existing machine process for a new product. This greatly reduces the initial startup costs and time when testing something new, thereby limiting the risk that’s often associated with new product development.
Machine vision also offers a lot more consistency than more manual processes. The visual information cameras can pick up is highly detailed—enabling machine vision systems to conduct in-depth part inspections and defect detection. This decreases the amount of waste and defects throughout manufacturing, saving time and critical resources while additionally minimizing the potential for customer returns down the line.
2D vs. 3D Vision
Machine vision setups offer more consistency and efficiency than automated processes that don’t use machine vision. While newer 3D cameras are increasing in prevalence, there are some cases where using 2D cameras is still preferable.
2D machine vision is both cheaper and more efficient. 2D is often used in combination with backlighting to enable quick and easy part identification. Because the 2D images are less complex than 3D, 2D systems are much faster than their 3D counterparts. This increases the overall throughput of the manufacturing process.
Additionally, by using 2D cameras at differing points in the manufacturing process, multiple 2D pictures can be integrated to resemble 3D camera output. Overall, 2D is an ideal fit for identifying and locating parts with a simple shape.
2D machine vision is also perfect for barcode reading throughout manufacturing processes. The use of barcodes and QR codes on individual product components has effectively transformed traceability metrics throughout supply chains. Companies can now use machine vision to track, trace, and report key supply chain metrics to stakeholders.
Where 3D vision can be helpful is for use with especially complex shapes. In the last decade, camera costs have greatly reduced, allowing for automated manufacturing processes to use higher-quality and therefore more accurate cameras. 3D vision systems are no exception. High-quality 3D cameras can capture helpful Z coordinate data that 2D cameras cannot.
Most automated systems on the market are now fitted with machine vision in some capacity.
Wauseon Machine is one such company, using machine vision on all of its machines. Wauseon Machine has also developed innovative solutions to process issues such as light exposure from outside sources during manufacturing. By incorporating infrared light in machine vision camera systems, the captured light is put out of the spectrum of visible light. This enables more accurate camera readings in the case of light infiltration from disruption around the assembly process.
Machine vision is revolutionizing automated manufacturing, replacing older manual processes with cameras that locate and position product parts accurately. It enhances efficiency and reduces alignment errors, using 2D or 3D cameras to inspect and track complex-shaped parts, as well as read barcodes and QR codes.
Machine vision streamlines operations, reducing startup costs, waste, defects, and customer returns. It offers consistency and in-depth inspections, saving time and resources. While 2D vision is cheaper and faster, suitable for simple shapes and barcode reading, 3D vision excels with complex shapes, capturing Z-coordinate data.
Machine vision is now a standard feature in automated systems, improving manufacturing processes across the board.