Automated Copper Alloy Grain Size Evaluation Using a Deep-learning CNN
Moog uses a variety of metallic and non-metallic materials in the products that it manufactures. Often these products are used in weight critical and/or safety critical applications. Material quality often matters a great deal and is checked by engineers and technicians working in production or in Moog’s Material & Process Engineering group. We also spend time at Moog fine tuning production processes using statistical methods such as a Design of Experiments. With the advent of usable Artificial Intelligence and more automation a group of engineers at Moog decided to try to improve upon current methods for the evaluation of a copper alloy. The paper titled “Automated Copper Alloy Grain Size Evaluation Using a Deep-learning CNN” is the outcome of this work. This is one of the first papers that we are aware of at the intersection of AI and industrial statistics and shows how convolutional neural networks might be set up for real commercial use.