Methods and apparatus for machine learning predictions of manufacturing processes
The subject technology is related to methods and apparatus for training a set of regression machine learning models with a training set to produce a set of predictive values for a pending manufacturing request, the training set including data extracted from a set of manufacturing transactions submitted by a set of entities of a supply chain. A multi-objective optimization model is implemented to (1) receive an input including the set of predictive values and a set of features of a physical object, and (2) generate an output with a set of attributes associated with a manufacture of the physical object in response to receiving the input, the output complying with a multi-objective condition satisfied in the multi-objective optimization model.
1. One or more non-transitory computer memory devices storing software instruction for controlling one or more processors to provide:
a user interface configured to receive, at the one or more processors, a manufacturing process request, wherein the manufacturing process request includes a digital model representative of a physical object; and
a predictive engine configured to generate at least one predictive value associated with the manufacturing process request, wherein the predictive engine is configured based at least in part upon the digital model or upon one or more features derived from the digital model, and based at least in part upon one or more machine learning models; and
wherein the user interface is also configured to provide a prediction report associated with manufacturing the physical object, the prediction report including information based upon the at least one predictive value.
2. The one or more non-transitory computer memory devices of claim 1 , wherein the one or more features include (i) values corresponding to surface orientations associated with the physical object and (ii) values corresponding to non-planar surface areas associated with the physical object.
3. The one or more non-transitory computer memory devices of claim 2 , wherein the one or more features further include one or more of the following: (iii) values corresponding to a volume of the physical object, (iv) values corresponding to a surface area of the physical object, (v) values corresponding to convex hull volume of the physical object, and (vi) values corresponding to length, width and height of a prism enclosing the physical object.
4. The one or more non-transitory computer memory devices of claim 1 , wherein at least one of the machine learning models is trained using unsupervised machine learning.
5. The one or more non-transitory computer memory devices of claim 1 , wherein at least one of the machine learning models is trained with an incomplete training set.
6. The one or more non-transitory computer memory devices of claim 1 , wherein the predictive engine also produces a deterministic property and the prediction report also includes information based upon the deterministic property.
7. The one or more non-transitory computer memory devices of claim 1 , wherein information based upon the predictive value includes probabilities associated with entities of a supply chain.
8. The one or more non-transitory computer memory devices of claim 1 , wherein the predictive engine further takes as input from the manufacturing process request, one or more of (a) a number of requested parts, (b) previous manufacturing transactions committed by a requesting entity, or (c) time when a request is expected to be fulfilled.
9. The one or more non-transitory computer memory devices of claim 1 , wherein the predictive engine produces the at least one predictive value based further upon one or more candidate shape distributions retrieved as a pre-classifier of physical object shape.
10. The one or more non-transitory computer memory devices of claim 1 , wherein at least one of the machine learning models is trained using unsupervised machine learning trained with an incomplete training set.
11. The one or more non-transitory computer memory devices of claim 1 , wherein the predictive value is in the form of a probability distribution function.
12. The one or more non-transitory computer memory devices of claim 1 , wherein the predictive value includes a probability that a request will be accepted by one or more entities of a supply chain.
13. The one or more non-transitory computer memory devices of claim 1 , wherein the predictive value includes a probability that a manufacturing process will be authorized by a requester entity.
14. The one or more non-transitory computer memory devices of claim 1 , wherein the manufacturing process request further includes, and the predictive engine further takes as input from the manufacturing process request, one or more of (a) a requested fabrication material, (b) a requested surface finish, or (c) a requested tolerance.
15. The one or more non-transitory computer memory devices of claim 1 , wherein the information based upon the at least one predictive value in the prediction report includes cost information.
16. The one or more non-transitory computer memory devices of claim 1 , wherein the information based upon the at least one predictive value in the prediction report includes a quote to manufacture one or more of the physical object.
17. The one or more non-transitory computer memory devices of claim 1 , wherein the prediction report is provided by the user interface in near real-time upon receiving the manufacturing process request by the user interface.
18. The one or more non-transitory computer memory devices of claim 1 , wherein the software instruction further provide an object analysis engine that generates from the digital model, a set of features associated with the physical object corresponding to at least one of (a) a discretized version of the digital model or (b) a point cloud processing of the digital model.
19. The one or more non-transitory computer memory devices of claim 1 , wherein the information complies with a multi-objective condition satisfied in a multi-objective optimization model.
20. One or more non-transitory computer memory devices storing software instruction for controlling one or more processors to provide:
a user interface configured to receive, at the one or more processors, a manufacturing process request, wherein the manufacturing process request includes a digital model representative of a physical object; and
a predictive engine configured to generate at least one predictive value associated with the manufacturing process request, wherein the predictive engine is configured based at least in part upon the digital model or one or more features derived from the digital model, and based at least in part upon one or more regression machine learning models; and
wherein the user interface is also configured to provide a prediction report associated with manufacturing the physical object, the prediction report including manufacturing cost information based upon the at least one predictive value.