A machine learning (ML) company Printpal, has launched a new artificial intelligence-based defect detection software that works with 3D printers – PrintWatch.
The plugin PrintWatch detects 3D printer build areas by utilizing a camera aimed at the build area. Using machine learning, the software detects exactly when a defect appears during 3D print jobs in real-time. Using PrintWatch, customers can track defects in the 3D printing process, allowing them to take proactive action if necessary.
Besides the ability to abort print jobs and turn off a 3D printer’s heat, Printpal’s new software also sends notifications about status alerts. Essentially, the software prevents printers from continuing to print a defective print for hours at a time, which saves time and filament while reducing the risk of hardware damage or a fire.
PrintWatch’s core technology is its computer vision capabilities. Through a proprietary machine learning model, the software can detect spaghetti-like defects in real-time, regardless of size, shape, colors, materials, lighting, or settings. Printpal is the first software to use extensive real-world data to train its detection model in dynamic settings, meaning the software will work with any FFF 3D printer.
The software tracks the progress of a defect to determine whether it is improving and only then intervenes if necessary. The tracking approach keeps false positives from occurring, so successful print runs should not be accidentally canceled.
A local running edition of Printpal is currently being developed so that it can be used on machines without an Internet connection. Offline versions are best suited to industrial printing operations that require high levels of network security.
Additionally, the company is developing several other AI-based 3D printer systems. There is software that optimizes the speed and quality of a build by analyzing the G-Code, inference APIs that detect objects using computer vision, and software that integrates e-commerce into 3D printers.