Automated defect detection – RnD project AM Intelligence presented at RAPID + TCT – Los Angeles

Interspectral has collaborated with AMEXCI in the RnD project AM Intelligence with the goal to build a platform for data driven defect detection in additive manufacturing. The result will now be presented at Rapid + TCT in Los Angeles, North America’s largest additive manufacturing event.

Since 2023, Interspectral has partnered with AMEXCI, a Swedish powerhouse within metal AM in the RnD project “AM Intelligence”.  The project has been granted 500000 Euros fund by the Swedish innovation agency Vinnova and have also included the German research institute Fraunhofer ILT.

“We had very high ambitions starting this project – the goal was not only to explore a new field but also to develop a solution that generates real user value and put it in the hands of AM end users. So, we are now very glad that we at RAPID + TCT can present a production-ready solution that is already running live at customer sites.” – says Isabelle Hachette, CEO at Interspectral.

Additive manufacturing is a rapidly growing field that has the potential to transform industrial production. However, one of the challenges the AM industry is facing is the ability to detect defects in 3D-printed parts. This is where AM Intelligence project comes in.

AM Intelligence is a project aimed at exploring how high-resolution sensor data, data fusion, advanced machine learning algorithms, and 3D visualization can support users to detect defects more efficiently in 3D-printed parts. This allows manufacturers to quickly and accurately identify and fix defects, improving the quality and reliability of their products.

The project has resulted in a modular solution for AI powered automated defect detection that will accelerate the workflow for quality assurance.  The solution has already been deployed successfully at customer sites.

“We are excited to see the impact of AM Intelligence on the industry. The collaboration with Interspectral and Fraunhofer ILT has significantly advanced our capabilities in defect detection for additive manufacturing.” – says Manja Franke, CSO at AMEXCI.

At RAPID + TCT Interspectral will present an AM Explorer demo that showcase an automated workflow for an EOS M290 machine, one of the most sold machines on the market. 

All process monitoring data is collected via EOSCONNECT and then AI powered analysis is performed on the powder bed images. The indications, including types, size and severity, are then presented for the end user in rich interactive visualizations together with process monitoring data.

The end user can then apply filters, look for root causes and then finally generate a report.

“The ability to visualize and analyze data interactively allows our users to quickly identify and address defects, significantly improving the efficiency and reliability of their additive manufacturing processes,”– says Isabelle Hachette, CEO of Interspectral.

For those interested in advancing their additive manufacturing operations, we invite you to visit Interspectral at booth #959, Hall G, during RAPID + TCT. Engage with our team to explore how our solutions can optimize your AM processes. To schedule a meeting with Interspectral at RAPID + TCT, please click here.

AM Explorer now includes AI powered automated powder bed image analysis. With the modular architecture end users can choose from off the shelf AI models or develop their own.