This technology makes it possible to read barcodes, measure fill levels, position robot grippers, check assembly and much more – machine vision is indispensable for manufacturing automation. We wanted to know why B&R vision systems are particularly well-suited for adaptive manufacturing and what added value they bring to applications. Our partners in this conversation are Nikolai Feurer, packaging technology expert at B&R, and Andreas Waldl, product manager for vision systems.
Andreas Waldl: A machine that can automatically adapt to continuously changing products and requirements must be able to "see". A vision system, which we have fully integrated into our control technology, serves as the eye.
Nikolai Feurer: Let's take a filling line for shampoo bottles of different shapes and sizes as an example. The empty bottles are fed onto a conveyor belt in any order. The bottles have to be placed in a vertical position before filling and pick-and-place robots are generally used for this purpose. To be able to grip the bottle, the controller needs real-time information about where the bottle is and how it is oriented. The dimensions are also required in order to clamp the bottles between two shuttles on the ACOPOStrak transport system. A crucial point is that all these processes rely on the data from the machine vision system and must be synchronized within the range of a few microseconds.
Waldl: Exactly. In adaptive manufacturing, image recognition needs to be able to deal with the fact that the product being recognized changes with each cycle – in the most extreme case as a batch size of 1.
The faster the lighting adapts to the respective product, the higher the clock rate. Due to its deep integration into the automation solution, the B&R vision system, and the lighting in particular, can flexibly adapt to changes during runtime without sacrificing recognition quality. Illumination with maximum homogeneity and intensity, achieved in as little as 150 nanoseconds, ensures that all recognition processes run at the highest possible clock rate with reproducible results.
Feurer: Our vision solution is fully integrated into the B&R system landscape, enabling microsecond communication with all the controllers, drives, safety technology and industrial PCs in the B&R portfolio. Now the camera can synchronize with the axis or use a motor position as a trigger. Especially for highly dynamic applications, I use it to create pinpoint images without lowering the output because of stopping or decelerating.
Waldl: The prerequisite for reliable recognition is product-specific lighting that adapts the lighting angle, light color and exposure duration in each case according to the situation. For this purpose, B&R offers a selection of external lighting systems that can be switched to meet specific requirements within a very short time, which is unique in the industry.
Waldl: Yes, it does. A customer opts for an adaptive machine solution because it allows mass customization, which refers to customized products with the economic conditions of mass production. Ultimately, the central consideration is therefore how economical is the chosen machine vision solution. So we're talking about the contribution of a machine vision solution to overall equipment effectiveness (OEE) and the cost of operating it over its complete lifecycle – in other words, the total cost of ownership (TCO).
Waldl: On the one hand, they save material and energy resulting from scrap that would otherwise occur due to poor or missing synchronization during frequent product changes. Malfunctions caused by extraneous light or optical interference can also be eliminated with the B&R vision system, thereby increasing total output. But overall, it is the fact that the adaptive system recognizes a wide variety of products at the highest clock rates with maximum reliability.
Feurer: The benefits extend across the entire lifecycle – from planning, project engineering and commissioning to operation and service. As a full-fledged station on the POWERLINK communication bus, the vision system is configured in the same way as any other component in Automation Studio using a its own mapp component – mapp Vision. This makes engineering, commissioning, maintenance, configuration, diagnostics or installing new firmware easier. In addition, the vision system is immediately integrated into the HMI application.
Waldl: It starts with engineering. It's much easier and less error-prone when the development team works in a common tool, such as Automation Studio, instead of two or more different tools.
In addition, the integration facilitates plug and play functionality. Out of the box, our vision sensors are not yet equipped with any functions, such as recognition of patterns or barcodes. When the vision sensor is connected to the machine, it automatically receives all the required functions and parameters from the control system. And if the operator needs to replace the camera or the lighting, reconfiguration is also done via plug and play. B&R cameras and lighting are factory calibrated. Reproducible results without adjustments to the exposure parameters on the machine are thus guaranteed.
Feurer: The bottom line is that you no longer need a vision expert for a standard application; you can configure applications using the intuitive mapp Vision interface.
Waldl: Yes, it still is. Traditionally, a vision system is a kind of "island" with its own programming. It is not integrated as an equal participant in the overall automation system and merely passes data to the bus via a standard interface. These are not seamless solutions and often only offer average performance in total because the control loops with the connected control system are not optimized. For the user, this means higher costs due to the additional work required.
Mr. Waldl, Mr. Feurer, thank you for your exciting insight into the world of machine vision.
