Inspecting a product made by additive manufacturing
Methods of inspecting a product made by additive manufacturing (AM) of multiple layers, computer program products and inspection modules for AM systems are provided. An augmented file is derived from a design file including layer data used to produce the product by AM. For each design layer, the augmented file includes the layer data for the design layer and weighted layer data for design layers beneath the design layer. A machine learning (ML) algorithm (trained on previous images and augmented files) is applied with respect to the derived augmented file onto received optical inspection images of the product during the AM process to detect production errors.
1. A method of inspecting a product made by additive manufacturing (AM) of multiple layers, the method comprising:
deriving an augmented file, from a design file including layer data used to produce the product by AM, wherein for each design layer, the augmented file includes the layer data for said design layer and weighted layer data for design layers beneath said design layer, and
applying a machine learning (ML) algorithm with respect to the derived augmented file—onto received optical inspection images of the product during the AM process to detect production errors,
wherein the ML algorithm is trained using a plurality of augmented files and corresponding optical inspection images—to adjust the weights so that production errors in each produced layer are detected irrespective of signals from the produced layers beneath said produced layer.
2. The method of claim 1 , wherein the ML algorithm is a deep learning algorithm.
3. The method of claim 1 , wherein the weights decrease with designed layer depth.
4. The method of claim 1 , wherein the augmented file includes the weighted layer data with decreasing darkness and/or contrast for design layers deeper below said design layer.
5. The method of claim 1 , wherein the detection of production errors includes detection of shifting of openings through two or more layers.
6. The method of claim 1 , wherein the product includes at least one semi-transparent portion.
7. The method of claim 1 , wherein the product is an electronic device that includes patterned conductive material and semi-transparent insulating material.
8. The method of claim 1 , wherein the product is a non-conductive semi-transparent model.
9. A computer program product comprising a non-transitory computer readable storage medium having computer readable program embodied therewith and configured to carry out at least partially the method of claim 1 .
10. A computer program product comprising a non-transitory computer readable storage medium having computer readable program embodied therewith, the computer readable program comprising:
computer readable program configured to derive an augmented file, from a design file including layer data used to produce a product made by additive manufacturing (AM) of multiple layers, wherein for each design layer, the augmented file includes the layer data for said design layer and weighted layer data for design layers beneath said design layer, and
computer readable program configured to apply a machine learning (ML) algorithm with respect to the derived augmented file—onto received optical inspection images of the product during the AM process to detect production errors,
wherein the ML algorithm is trained using a plurality of augmented files and corresponding optical inspection images—to adjust the weights so that production errors in each produced layer are detected irrespective of signals from the produced layers beneath said produced layer.
11. A computer program product comprising a non-transitory computer readable storage medium having computer readable program embodied therewith, the computer readable program comprising:
computer readable program configured to apply a machine learning (ML) algorithm with respect to a derived augmented file, the augmented file derived from a design file including layer data used to produce a product made by additive manufacturing (AM) of multiple layers, wherein for each design layer, the augmented file includes the layer data for said design layer and weighted layer data for design layers beneath said design layer—onto received optical inspection images of the product during the AM process to detect production errors,
wherein the ML algorithm is trained using a plurality of augmented files and corresponding optical inspection images—to adjust the weights so that production errors in each produced layer are detected irrespective of signals from the produced layers beneath said produced layer.
12. An inspection module in an additive manufacturing (AM) system, the inspection module comprising the computer program product of claim 11 .