Development of a parametric production jacket pattern for an automated pattern-making system

Kim, N. R., & Park, J.H. (2025). Development of a parametric production jacket pattern for an automated pattern-making system.” Fashion and Textiles 12.1 (2025): 1-17

Abstract

This study presents a parametric production jacket pattern–generation method designed to automate the pattern-making process. Building upon a previously developed parametric nude pattern—constructed from radial body-surface measurements of a central-size 3D body model—the research integrates industrial ease allowances and essential jacket design elements to establish drafting formulas applicable across multiple sizes. The resulting parametric production patterns were validated against conventional graded patterns, demonstrating high similarity in silhouette and structural lines and confirming the potential for algorithmic automation in ready-made apparel patternmaking.

Purpose

The purpose of this study is to develop a parametric production jacket pattern that can be automatically generated using only two key body measurements: height and bust circumference. Traditional production patterns rely on tacit knowledge, fitting iterations, and manual grading, limiting opportunities for automation. This research aims to:

• Transform the parametric nude pattern into a production-ready jacket pattern.

• Integrate ease allowances and fundamental design elements (e.g., princess-line division, dart positions, dart amounts).

Verify whether parametric production patterns can reliably replace conventional graded patterns.

Method

1) Construction of Parametric Nude Pattern

Radial body-surface lengths from the bust and scapular points of a central size dummy were used to derive proportional drafting formulas. These formulas generated zero-ease nude patterns for the central, upper, and lower size categories. For jacket application, hip circumference and hip length were also incorporated as proportional variables.

2) Generation of Parametric Basic Pattern

Ease allowances and baseline adjustments—such as armhole depth—were extracted by overlaying the nude pattern with an industrial production pattern in YUKA CAD. Differences in line lengths and heights were converted into constants and added to the proportional equations. The resulting formulas enabled consistent pattern generation based solely on height and bust circumference.

3) Development of Parametric Production Pattern

Design elements including princess-line divisions, dart centers, dart lengths, and dart amounts were analyzed across three sizes. Positional changes were expressed as proportional ratios, while dart amounts were defined as constants for production efficiency. These rules were integrated into the drafting system to generate production patterns for three size categories without using conventional grading.

4) Verification

Parametric patterns were compared to graded patterns via dimension analysis and overlay. Overall seam lines showed high similarity, especially in the central size, although deviations appeared in areas with high curvature (front dart region, waistline, armhole, and back shoulder slope). Relative errors of 5–15% in the front-panel dart confirmed the need for further refinement in curved regions.

Discussion & Conclusion

This study proposes a viable framework for parametric pattern automation, showing that jacket production patterns can be generated using only two variables while preserving the logic of manual drafting. The parametric nude pattern also demonstrates applicability to tight-fitting wearable products without algorithmic modification.

However, limitations remain:

• discrepancies in curved regions (shoulder slope, front darts),

• restricted garment category (jackets only),

• limited size range,

• insufficient representation of real body-shape diversity.

  
  

Future work should expand to various garment types, broader size ranges, and more diverse anthropometric datasets. The study also highlights the distinction between mass production–focused automation (the scope of this research) and mass customization (e.g., body-type pattern libraries), suggesting that each requires tailored algorithmic strategies.

While using only height and bust circumference simplifies automation, the sufficiency of two variables for industrial replacement remains debatable. Increasing the number of variables may improve accuracy but risks reducing computational efficiency. Thus, determining the minimal yet adequate set of variables is essential for advancing automated pattern generation.