<p>The modular design maximizes utility by employing standardized rather than customized components in large-scale systems. From a manufacturing view, this approach contributes to green technology by reducing material waste and enhancing component reusability. Additionally, from an industrial view, it offers significant economic advantages by leveraging economies of scale, making it a highly reasonable strategy. Modularization is generally achieved by selecting a representative design that satisfies all required performance from a predefined set of designs. However, achieving effective modularization in mechanical mechanism systems presents additional challenges. First, mechanisms depend on geometric relationships, and varying loads yield different optimal designs. Second, selecting a single design to cover all conditions for modularization inherently results in an over-specified solution. This results in inevitable performance deviations, which become pronounced as the design scale increases. To address these challenges, a modular mechanism design framework based on surrogate-based design optimization is proposed. Surrogate-based optimization is utilized to obtain optimal design solutions for each condition in large-scale engineering systems. The generated optimal designs are partitioned into multiple groups, each unified by a representative design covering all its conditions. Thus, it leads to a multi-objective optimization (MOO) problem that aims to maximize economies of scale while minimizing performance deviations. Unlike conventional approaches, which primarily rely on intuitive design candidates and simple grouping strategies, the proposed framework generates optimal solutions directly, enabling the selection of optimal layouts. Additionally, it analyzes manufacturing cost parameters to support strategy selection for various design scenarios. This approach enhances maintainability and promotes environmentally sustainable manufacturing.</p>