Digital Twin-Driven Supply Chain Enhancement to Support Direct-to-Consumer Growth

Abstract

In response to the rising trend of Direct-to-Consumer (D2C) sales, many traditional retailers, which have historically relied on wholesale business models, are now undertaking significant supply chain transformations. This thesis explores the strategic shift of a large retailer in the footwear and apparel sector, pseudonymously referred to as Iota in this study, as it transitions towards a D2C-focused supply chain. This transition, emblematic of a broader industry transformation, is aimed at enhancing alignment with the evolving expectations of customers in terms of service, cost-effectiveness, and sustainability.

Central to this research are the proposed enhancements by Iota’s leadership to decentralize Iota’s supply chain. These enhancements include adding both physical infrastructure, with the planned establishment of a cross-dock facility, and digital infrastructure, through the development of a decision engine that aids in efficiently routing products within the new decentralized supply chain network. The cross-dock facility is envisioned to provide an opportunity for decision postponement in the inventory flow from Asian factories to US distribution centers. Meanwhile, the decision engine, leveraging a heuristic-based algorithm, is set to unlock new inventory flows and enhance inventory distribution.

With the new infrastructure to decentralize the supply chain yet to be fully operational, a retrospective study was conducted using a digital twin of Iota’s supply chain. Various push and pull-based inventory deployment strategies were simulated in the digital twin with the goal of alleviating pressure on the primary distribution center and increasing fulfillment from regional distribution centers. In the simulation process, challenges with forecast data and lumpiness of supply are discovered and subsequently addressed through the use of synthetic datasets, which emulate improved forecast coverage and smooth supply.

The key findings from simulations highlight that despite achieving a modest performance in meeting the goals for the decentralized network, valuable insights were obtained that could drive future supply chain enhancements. The research underscores the benefits of smoothing supply for network performance, the critical role of comprehensive and reliable forecast data, and the necessity for supplementary storage solutions to complement the cross-dock facility. For example, one pull-based scenario using a synthetic dataset to emulate enhanced forecast coverage and smoother supply tripled network performance while reducing network costs by 1% compared to the baseline pull-based scenario. Such cost savings could be substantial for a large- scale retailer.

Concluding with recommendations, the thesis advises Iota to re-evaluate purchasing practices, consider integrating multiple internal sources of forecast data into a single source, and continue with simulation analyses. These recommendations are designed to support Iota, and by extension, similar retailers, in their transition towards a robust and agile D2C supply chain, ensuring competitive advantage in the dynamic retail sector.