Superficially Smart Artificial Intelligence Detects Errors in Painting

Source: Fraunhofer-IPA | Translated by AI 2 min Reading Time

Related Vendor

"Intelligent" algorithms could, according to Fraunhofer researchers, soon warn of painting errors and machine failures before it is too late.

Painting test with data collection at Fraunhofer-IPA! A robot sprays paint onto a plastic car bumper. An inline laser detector and other sensors record everything. The data was then used to train artificial intelligence (AI). There’s a reason for this ...(Image: Fraunhofer IPA / R. Bez)
Painting test with data collection at Fraunhofer-IPA! A robot sprays paint onto a plastic car bumper. An inline laser detector and other sensors record everything. The data was then used to train artificial intelligence (AI). There’s a reason for this ...
(Image: Fraunhofer IPA / R. Bez)

Painting, the IPA researchers note, is still considered a process that cannot be fully controlled. This means there is still a risk of waste, equipment failures, and rework, as the specified paint layer thickness is very often not achieved everywhere. At the Fraunhofer Institute for Manufacturing Engineering and Automation (IPA), the goal of a research project is to use AI to reduce the number of errors in the painting process and machine downtime. The project is called "pAInt-Behaviour" (Efficiency improvement of painting processes through multilayer networking of process and quality data using self-learning behavioral modules).

A Car Bumper As An Object for Optimization

An IPA research team has now conducted a series of tests focusing on the painting of plastic parts. Specifically, it involved a car bumper fixed in a frame wrapped with aluminum foil. The data collected was then used by the experts to train artificial intelligence (AI), it was further stated. And this is how the process can be imagined: From the left, a robot arm moves towards the component. Once it has reached its starting position, it pauses briefly before proceeding to paint the plastic part. The robot repeatedly interrupts its work briefly and changes position to paint the bumper evenly from all sides. Additionally, the settings were changed before each individual test! This was done intentionally to induce errors. But the unique aspect of this system, located in the coating technology center at IPA, largely remains hidden to the human eye! This is because the robot-assisted painting process is monitored by artificial intelligence (AI).

This is How Plastic Parts Are Smartly Painted

The most noticeable aspect might be the inline laser detector from AOM, mounted above the robot's spray nozzle. During the painting process, it captures the number, size, and speed of the paint droplets. This data is then fed into a database where, as emphasized, all measurement and process data converge, supplied by 30 different sensors. These include data from the "b+m" system control, such as rotational speed, voltage, valve switching, and the amount of paint consumed. It also includes the amount of air directing the paint droplets, to name just a few parameters. Added to this are the results of measurements taken on the finished painted bumper, along with those from a visual inspection by an experienced paint master (paint layer thickness by experts from Helmut Fischer, color tone, gloss, waviness, and foreign particle inclusions).

Subscribe to the newsletter now

Don't Miss out on Our Best Content

By clicking on „Subscribe to Newsletter“ I agree to the processing and use of my data according to the consent form (please expand for details) and accept the Terms of Use. For more information, please see our Privacy Policy. The consent declaration relates, among other things, to the sending of editorial newsletters by email and to data matching for marketing purposes with selected advertising partners (e.g., LinkedIn, Google, Meta)

Unfold for details of your consent