US 12,422,826 B2
Systems and methods for generating a zero-waste design pattern and reduction in material waste
Shelly Xu, Chatham Township, NJ (US); and Tuanfeng Wang, Chatham Township, NJ (US)
Assigned to SXD, INC., Chatham Township (JE)
Filed by SXD, Inc., Chatham Township, NJ (US)
Filed on May 4, 2023, as Appl. No. 18/143,073.
Claims priority of provisional application 63/340,684, filed on May 11, 2022.
Prior Publication US 2023/0367292 A1, Nov. 16, 2023
Int. Cl. G05B 19/4155 (2006.01); A41H 3/00 (2006.01); G06Q 10/04 (2023.01)
CPC G05B 19/4155 (2013.01) [A41H 3/007 (2013.01); G06Q 10/043 (2013.01); Y02P 80/40 (2015.11)] 6 Claims
OG exemplary drawing
 
1. A multi-task optimization process for generating a zero-waste design pattern and reduction in waste of a source fabric/material, comprising a first process comprising the steps of:
a) providing one or more inputs required for material fabrication to a system (100), wherein the provided one or more inputs include material information (401), a plurality of cut pieces (402) that are either taken directly from a pattern or a tech pack (403) or derived from a 2D/3D design sketch (404), and a metadata indicating one or more characteristics required for the material fabrication;
b) retrieving a template or a combination of template parts from a template library (405) based on one or more inputs, wherein the template contains a set of default fabric/material pieces that correlate with the metadata;
c) scaling the set of default fabric/material pieces that correlate with the metadata to the pattern or the tech pack (403) or the 2D/3D design sketch (404);
d) generating a first 3D surface (406) by rendering an assembled garment on a virtual mannequin using the material information (401) and the cut pieces (402), wherein the first 3D surface (406) provides a perceptual constraint on the multi-task optimization of the pattern;
e) comparing the first 3D surface (406) with a second 3D surface stemming from the pattern or the tech pack (403) or the 2D/3D design sketch (404) to determine the distortion and iterating the comparison, until distortion between the first 3D surface (406) and the second 3D surface is reduced to a minimum threshold value;
f) checking one or more pairs of neighboring patches on the first 3D surface (406), wherein a patch merge (407) for a pair of patches, each patch merge forming a material patch piece, is performed if:
i) a curvature across at least one sewing edge is less than a predetermined threshold; and
ii) the merging of two of the at least one sewing edge are small enough to fit within a dimension of the first 3D surface (406);
g) performing patch shape optimization (408) over at least one material patch piece, wherein each material patch piece is represented as a set of pre-defined shapes assembled to reflect curvature of the material patch piece according to the first 3D surface (406);
h) identifying the curvature of each material patch piece by splitting the material patch piece along at least one boundary having a sharp connection, wherein for each adjacent pair of the curvature of the material patch piece, each boundary is down sampled by gradually reducing the number of vertices along each boundary;
i) performing a 2D strip packing (409) to optimize a shape of the material patch piece, wherein a minimum area of a bounding box is computed for each material patch piece after packing; and
j) computing an empty area for each bounding box and selecting the patches whose empty area exceeds a predetermined threshold value, wherein patch splitting (410) is performed for the selected patches which are divided into smaller patches to achieve maximum space optimization.