A build engineer comes in on Monday to find a laser powder bed fusion job finished over the weekend. The parts look fine. Two weeks later the CT results come back showing a warped internal geometry, or a streak of poorly melted material that started somewhere around layer 400. Much of what printed after that point, along with the machine hours, the powder and the testing fees, may end up as scrap.
That gap between cause and discovery is one of the less visible obstacles to scaling metal LPBF into serial production. Defects such as warpage, powder short feed and streaking are frequently caught only in post-processing or testing, days or weeks after a build finishes. In-process monitoring can help, but it often means retrofitting machines with dedicated sensors, which is manageable for a couple of printers and harder to justify across a fleet.
Icelandic software startup Euler takes a different route. Its platform is software only and hardware agnostic, built on the observation that most modern LPBF systems already capture layer images during every build, and that in many cases those gigabytes of data sit unused on a hard drive. Because the camera is already in place, the company says it can connect to an existing setup in under five minutes, with results visible immediately.
Rather than flagging defects after they occur, when it is already too late, Euler takes a predictive approach that alerts the operator well in advance. Most defects show early indicators, and Euler’s patent-pending predictive engine is built to detect them, warning of a likely failure before a job is lost.
Study quantifies the cost of failed builds
Published evidence for the approach comes from a joint study with the Danish Technological Institute (DTI). Over 47 days, Euler monitored 54 builds across four Nikon SLM systems at DTI, covering six metal alloys including 316L stainless steel and Ti6Al4V titanium.
The paper found that 27.8% of total print time was spent on builds that ultimately failed. Of that wasted time, 77% could have been saved by halting at the point of Euler’s failure prediction, equivalent to roughly 10 days of 3D printing.
Referring to a Wohlers Report’s estimated operating cost of $75 per hour for a $1M machine, the authors put the potential annual saving at over $115,000 per machine in operating costs alone, before accounting for wasted powder, personnel time or post-processing.
One DTI case stood out. A production build was uploaded to Euler retrospectively, as part of a quality investigation. Manual review of nearly 14,000 layer images had failed to locate the problem.
Euler’s platform flagged the defective layers immediately and showed that an early alert would have saved 38 hours and 43 minutes of partial-part printing from becoming scrap and prevented a further 48 hours of wasted printing after the failure, amounting to over 86 hours of print time affected by the failure.

New integrations take centre stage at Formnext
For regulated industries, deployment matters alongside detection. Euler offers on-premises and private cloud options, including GovCloud, which can matter to aerospace and defense users who are unable to send build data to a public service.
Euler supports major metal printer brands, among them EOS, Nikon SLM and Renishaw, and collaborates with Autodesk and Scanlab. Frumtak and Kvanted, investors focused on deep tech and manufacturing technology, back the company.
Euler will be at Formnext in Frankfurt from November 17 to 20, in Hall 11.0 at booth A32, where the team will demo new connectivity features including Model Context Protocol integration and Slack and Microsoft Teams alerts. MCP lets engineers query their build data directly from AI assistants such as Claude or ChatGPT, asking questions like “Which builds this week showed defects, and on which machines?” and getting an answer in seconds.If you’re running a fleet and want to know which parts have defects before you ship them for testing, book a meeting here.
Can’t make it to Frankfurt, or want to see results now? Euler’s free Powder Bed Image Viewer lets you monitor builds from any browser, no remote-desktop workaround needed. Sign up here and have your first results in under five minutes.
To stay up to date with the latest 3D printing news, don’t forget to subscribe to the 3D Printing Industry newsletter or follow us on LinkedIn.
Explore the full Future of 3D Printing and Executive Survey series from 3D Printing Industry, featuring perspectives from CEOs, engineers, and industry leaders on the industrialization of additive manufacturing, 3D printing industry trends 2026, qualification, supply chains, and additive manufacturing industry analysis.
Featured image shows 3D visualization of an anomaly distribution (spatter) on a reconstructed 3D component. Image via Euler.






