In today's dynamic market environment, solving large-scale supply chain network design problems within a practical timeframe poses a significant computational challenge. Given the massive scale of real-world data, Mixed-Integer Linear Programming (MILP) models often fail to provide solutions quickly and may even encounter out-of-memory errors. To address this challenge, this study develops a systematic simplification framework that renders large-scale supply chain network design problems computationally tractable while largely preserving solution quality. The framework simplifies the input data through product aggregation and customer clustering. To validate that this simplification does not compromise solution quality, we apply the same MILP model to both the original and simplified versions of small-scale datasets and compare the outputs. The results confirm that our simplification framework works effectively, thereby allowing us to confidently apply it to solve large-scale, real-world supply chain network design problems. This framework offers practitioners a computationally efficient tool for rapid, data-driven network planning decisions.
