| CPC G06F 30/20 (2020.01) [G06N 5/02 (2013.01); G06N 20/00 (2019.01); G06T 11/20 (2013.01); G06T 2210/12 (2013.01)] | 14 Claims |
|
1. A method comprising:
by a computing system:
accessing a computer-aided design (CAD) assembly that digitally represents a product component that includes multiple parts;
performing a machine-learning (ML)-based part determination process for the CAD assembly, including by:
providing the CAD assembly as an input to a ML model trained with assembly structure data of CAD assemblies of a common product type as the CAD assembly;
generating a candidate part set through the ML model, the candidate part set comprising a plurality of candidate parts as candidates to include in the CAD assembly through the ML-based part determination process;
filtering the candidate part set based on geometric constraints of an assembly space at which a recommended part determined through the ML-based part determination process is to be positioned in the CAD assembly, including by:
performing a geometric analysis of the assembly space in the CAD assembly to determine a bounding box that specifies physical dimension constraints for the recommended part determined through the ML-based part determination process; and
removing, from the candidate part set, any candidate part of the plurality of candidate parts with part dimensions that exceed the bounding box; and
filtering the candidate part set based on physical and cost characteristics of the different candidate parts of the candidate part set, including by removing, from the candidate part set, any candidate part of the plurality of candidate parts with cost characteristics or physical characteristics that do not satisfy a filtering criterion; and
identifying the recommended part for inclusion in the CAD assembly from the filtered candidate part set; and
inserting the recommended part into the CAD assembly; and
providing the CAD assembly in support of physical manufacture of the product component represented by the CAD assembly.
|