Update

nebumind nominated for TCT awards

27/05/2024

The entries were judged by over twenty independent industry experts, and we are excited to be one of eight award finalists in the “software” category with our AI recoater eye.

Software innovation

nebumind has integrated its “AI Recoater Eye” in its digital twin software to automate the identification and reporting of defects during the application of a new powder layer. In the past, engineers had to evaluate powder bed images in a laborious manual process post-print. Now nebumind automatically detects defects during the printing process through AI-based image analysis and visualizes them in a digital twin of the component. Printer operators see this as a great time saver in terms of analysis efforts and as a significant cost saver in terms of rejects due to in-process quality monitoring and early countermeasures.

Uniqueness and originality this innovation brings

Unlike previous image recognition attempts, nebumind together with its pilot customer Isar Aerospace and its development partner EOS has successfully integrated the first App of its kind into serial production, enhanced with several unique features:

• 3D Digital Twin to speed up anomaly analysis: Anomalies are shown across multiple coating layers within a 3D digital twin of the actual component, enabling more precise assessment of the size and position of anomalies inside and outside components.
• In-Process Reporting to reduce waste: Anomalies are reported in real-time during the printing process, allowing for immediate countermeasures to avoid unnecessary printing time and rejects.
• Long-Term Monitoring to prevent recurring defects: Anomalies are tracked over several print jobs and individual components to spot recurring defects happening in the same layer and at the same position.
• App Transferability to avoid isolated solution: “AI Recoater Eye” is compatible with various printer types and automatically adjusts the sensitivity of its image recognition algorithm to changing factors such as material, camera type, and layer thickness through transfer learning, eliminating the huge manual effort to calibrate with changing environmental factors.

Impact this innovation will have

In-process quality monitoring is a critical element, but a blank spot in the 3D printing industry, with very few limited tools available on the market. nebumind’s “AI Recoater Eye” enables the user to take first steps towards the ultimate goal of an online monitoring capability. By automating quality analysis of coating images and moving it into the printing process, the App significantly drives down printing costs, increases the return on investment (ROI) and thus makes 3D printing more attractive to a broader range of companies. In essence, nebumind’s pioneering App not only addresses a critical deficiency, but also reshapes the economics of 3D printing, making it a more financially viable and an attractive option for a diverse array of industry players.

Environmental impact

The high scrap rate in 3D printing and the detection of defects only after a print job leads to a very negative CO2 footprint of 3D printed components. On the one hand, more energy is needed to produce components that meet minimum quality standards, on the other hand, more material is needed, which also requires a high CO2 input. By integrating real-time defect detection into the printing process, “AI Recoater Eye” dramatically reduces the likelihood of producing unusable components, thus lowering the energy and material waste associated with reprints and corrections.

Scope for further development of this innovation

To further automate the process, nebumind plans two extensions: On the one hand, nebumind and EOS work together in a public funded project to expand in-process countermeasures in case of a found coating anomaly, such as automatically applying a new coating layer or the possibility to take out defective components before continuing the print job. On the other hand, nebumind and Isar Aerospace plan to correlate found coating defects with other process data from the print job and CT data from the final quality check to identify root causes and prevent coating defects from happening in the first place.

In the long term, nebumind not only wants to monitor the printing process, but also actively regulate it. The goal is to establish a fully self-regulating manufacturing process that optimizes itself in real time, minimizing human intervention and maximizing production efficiency.

The TCT Awards Ceremony choosing the final winners takes place on the 5th June in Birmingham, UK. For more information go to TCT Awards | The World’s Leading 3D Printing & Additive Manufacturing Awards