by Melissa Donovan
Three-dimensional (3D) printing is a complex process that often involves many parts. If one step in the procedure goes wrong the entire build is jeopardized. Real-time monitoring and issue detection software are important tools to prevent this from occurring. Their use minimizes waste, saves time, and ultimately helps create an efficient method of manufacturing.
Above: Phase3D Fringe identifies anomalies within the build process.
What Is It
Real-time monitoring and issue detection software gather data during the 3D printing process. “The number of data sources is very broad, ranging from printing parameters, oxygen level sensors, and laser paths to more complex data sets like meltpool data, optical tomography, in-situ non-destructive testing, and layer images,” says Jan Van Espen, data intelligence manager, Materialise.
“Software uses a live camera feed to detect 3D printing failures in real time, and a connection to a printer through an external board, like a raspberry pi, to act when a failure occurs. This provides users with the security of running their printers without having to physically monitor them,” shares Peter Lebiedzinski, founder/CEO, printpal.io.
Real-time monitoring acts as a first defense. “It typically focuses on keeping an eye on whether machine parameters are within preset tolerances and raise alerts if not,” according to Van Espen.
On the off chance real-time monitoring fails, “issue detection software dives into a deeper level by identifying root causes,” adds Van Espen.
According to Mirco Schöpf, senior software manager, EOS, real-time monitoring and issue detection software is “typically a combination of hardware sensors of different wavelength measuring meltpool emissions and a software performing data acquisition, analysis, and visualization.”
Specifically in terms of powder-based additive manufacturing (AM), solutions like Phase3D’s Fringe measure the surface topology of the powder and melted and bound surface to identify anomalies within the build process. “The technology creates objective measurements of the build process, identifying how defective parts are produced and notifying the user early to stop the build. Fringe both measures the surface topology as well as true layer thickness, providing objective measurements to quantify and validate the build process for critical use parts,” shares Dr. Niall O’Dowd, founder/CEO, Phase3D.
Why It’s Important
All 3D printer users benefit from real-time monitoring and issue detection software, but some more than others.
“Real-time or in-situ monitoring stands as a pivotal technology capable of significantly curtailing costs and slashing build times for laser powder bed fusion AM processes. This holds immense significance not only for the AM part makers but also for the wider industrial landscape. Projections hint at a potential savings that could ascend to $150,000 per machine, annually,” notes Shuchi “SK” Khurana, founder/president/CEO, Addiguru.
Real-time monitoring is essential because it removes assumptions from the 3D manufacturing process. “Most manufacturers print parts and do not know the quality of the part for one to 12 months after the print is complete. With in-situ monitoring, the user can know in real time the quality of each part. This allows the user to stop a build early when a part is defective, which decreases project lead times and increases machine utilization while also ensuring part quality in real time,” shares O’Dowd.
Two industries that find real-time monitoring advantageous, aerospace and medical. Both have high requirements for quality control, according Van Espen. “The need for these solutions does not depend on the production volume but on the importance of guaranteeing quality, reducing scrap, and providing consistent production output.”
“Real-time monitoring is a simple and inexpensive way to prevent the financial and mental strain that sometimes comes with being a 3D printer user. I would recommend that all users either have a printer equipped with monitoring software or use a third-party monitoring software on their printers,” says Lebiedzinski.
Walking through an example, Lebiedzinski shares how many 3D printing jobs can take over ten hours to complete. “If the print fails two hours into the job without being detected, the printer will continue to print for the remaining eight hours, wasting time, material, energy, and potentially severely damaging the machine. Take for example a car, it tells you when it runs out of gas. Every home comes with fire alarms in case of a fire. Now take 3D printing, where failure rates can run as high as 60 percent, you would think that print monitoring would be a standard—if not required—offering.”
Standalone and Encompassing
Real-time monitoring and issue detection software are available as separate, standalone solutions or part of larger, all-encompassing software.
Standalone solutions are available that integrate into other software. For example, Phase3D’s Fringe is offered as a standalone solution with the capabilities to be incorporated into quality and process control and manufacturing execution systems, says O’Dowd.
“Standalone software solutions are available. However, to use them most efficiently, companies want solutions integrated into their workflow and production environment. As the quality of a final 3D printed part often depends on its post processing, manufacturers always need to be able to look at the production process in a true holistic way,” admits Van Espen.
Lebiedzinski says “typically, real-time monitoring is a feature provided by software companies that offer a full service cloud management solution and specialize in printer management tools.” However, printpal.io provides both a standalone monitoring solution and a partnership with cloud management provider, SimplyPrint, to serve consumers interested in a full service or standalone offering.
“A standalone offering is better suited for OEMs who may already have cloud management capabilities and are interested in integrating real-time monitoring to benefit their end consumers. A full-service offering is more suited to OEMs without either capability and for end consumers who want to be able to virtually manage their printers with the comfort provided by a real-time monitoring solution,” suggests Lebiedzinski.
Depending on the 3D printer manufacturer, real-time monitoring and issue detection software may be bundled into their solution. Take for an example EOS. “We offer fully integrated solutions with APIs to give partners and customers the opportunity to do their own analysis and draw their own conclusions,” shares Schöpf.
“A few 3D printer OEMs have recently started to implement monitoring solutions into their printers. Traditionally, the offering of real-time monitoring has been dominated by third-party cloud management providers who have built solutions on top of platforms like Klipper, OctoPrint, and RepRap/Duet3D. As the industry has grown and more consumers have begun to demand quality of life features—such as real-time monitoring—more OEMs have begun to explore offering defect detection,” explains Lebiedzinski.
The World of AI
Artificial intelligence (AI) plays a role in so many software solutions today, it shouldn’t come as a surprise that real-time monitoring and issue detection software is affected by AI.
Playing a large role in monitoring AM from machine learning techniques to analyzing build images, O’Dowd also believes AI is a tool to correlate build anomalies to defects, yet not to identify the anomalies themselves.
Lebiedzinski argues that you could not have an effective issue detection software solution without AI. “There are many variables in play inside a 3D printer, making it difficult for a human to model the ‘normal’ behavior of a 3D printer or any of its components. AI is excellent at dealing with this large amount of data. Most defects that occur in 3D printing are detected using a camera, which translates the images into a large amount of complex data. Dealing with this sort of data requires a comprehensive machine learning model that is trained on extensive datasets from thousands of printers using various cameras, lighting conditions, and camera angles.”
“AI enables insights into the 3D printing process by making correlations that are practically unfeasible to humans. It executes analytics quickly on huge data sizes and allows process and quality control to continue around the clock. Using AI also allows for the automation of tasks, effectively standardizing analytics and reducing the need for manual labor and the likelihood of human error. Instead of reviewing thousands of layer images of a 3D print job, experts can focus on tasks that add real value,” says Van Espen.
Integral Software
Real-time monitoring and issue detection software offer features that anyone tasked with managing 3D builds should take advantage of. Many printer OEMs provide variations of these solutions with their hardware, and there are also third-party options available that integrate with these products.
Apr2024, Industrial Print Magazine
