Artificial Intelligence to improve the quality of Additive Manufactured components

Additive Manufacturing (AM) is "a process of joining materials to make objects from 3D model data, usually layer upon layer, as opposed to subtractive manufacturing methodologies". One of the processes of AM is direct energy Deposition (DED). In the DED, a part is built by melting a surface with a laser beam and simultaneously applying metal wire or powder. The welding process is monitored by tuning robot control parameters, for instance, the distance of nozzle in relation to the substrate, the wire feed rate, or the laser power. By monitoring the melt pool with a camera, the images later will be used to analyze the deviations in welding process (good or bad weld images, below figure).

This is a very time-consuming verification procedure if done manually since there are a huge number of images generated from every built component, and prone to human-inaccuracy due to the time consuming and tedious task. Another problem is that material and time is wasted because the defect analysis is done after the component is manufactured. For instance, if larger defects are detected in the manufactured component, the product may be discarded. Here, artificial intelligent (AI) and prediction modelling come to play a role to automate the manual analysis using the process data. This helps to do both pre-analysis and post-analysis of in-situ monitoring to speed-up the criteria for the defect analysis.

GKN Aerospace have started to explore different AI methods to predict defects using image data from the DED processes. One of the challenges that we faced so far is that we need more labelling data to get more accurate models, and that another AI methodology for clustering could help in this process. From initial investigations, we see a great potential of using AI for both online monitoring of the process and as a method for reducing the cost of following nondestructive testing methods.