Using AI and Machine Vision to Improve Welding and Metal Additive Manufacturing Quality

Xiris combines AI and Machine Vision to deliver real-time, quantifiable insights for researchers, OEMs, and manufacturers advancing the future of weld quality.

Inside the Whitepaper

  • Using AI Feature and Object Detection in Welding

  • How AI and Machine Vision enable real-time anomaly detection and closed-loop control.

  • What indicators like melt pool geometry and cooling time reveal about weld quality.

  • How Xiris AI tools, such as MeltPool AI and Object Detection, are reshaping process monitoring.

  • Why HDR Visible and SWIR (Thermal) imaging give engineers data the human eye can’t capture.

  • Plus, the business case for welding with AI Vision, showing measurable ROI through scrap reduction, faster cycle times, improved safety, and full traceability.
W1 Asset 4 —Early access Email
Cooling Time GIF - Thermal Camera
Cooling Time GIF - Thermal Camera (2)
Cooling Time GIF - Thermal Camera (3)

Cooling Time in Metal AM

Metal AM involves building parts layer by layer, and controlling cooling rates is crucial for:

  • Achieving uniform mechanical properties throughout the build.

  • Avoiding defects caused by uneven cooling.

Cooling time measurement applies to both wire-based and powder-based processes, including:

  • Wire-Based Processes: GMAW-based WAAM, Plasma-Arc-based WAAM, Wire Laser AM (WLAM), Wire-based Electron Beam AM.

  • Powder-Based Processes: Laser DED, Laser Powder Bed Fusion, Electron Beam AM.

For wire-based methods like WAAM and WLAM, cooling time measurement is particularly critical. Due to the layer-by-layer nature of these processes, cooling rates can vary significantly within a single build, making accurate monitoring essential.