A simulation optimization framework for production planning and control in the fashion industry

Fashion is one of the world’s most important industries, driving a significant part of the global economy representing, if it were a country, the seventh-largest Gross Domestic Product (GDP) in the world in terms of market size. According to the high complexity that has to be managed by companies operating in the fashion Supply Chain (SC), Production Planning and Control (PP&C) represents a relevant issue that these companies have to face with, especially considering the dynamic context where they work and, consequently, the high occurrence of stochastic events (e.g. unexpected sample production, changes in production priority, raw material arrivals delay, rush orders) they have to manage. Even if this is a well-debated topic both from an academic and an industrial point of view, related tools are no widely adopted along companies working along the fashion SC, especially considering Small and Medium Enterprises (SMEs). The few implementations mainly refer to their adoption by brand owners for a single-step planning on a strategic and tactical level, while SMEs are discouraged because of the related complexity and high costs. According to this, the present work aims to present an iterative simulation-optimization framework for the fashion SC industry to be used by all the actors of the SC, both brand owners and suppliers, in order to continuously control, reallocate and optimize the production plan, considering their Critical Success Factors (CSFs) and the unexpected events that may occur. The reason why optimization and simulation are jointly used within this framework is twofold: on the one hand, using optimization algorithms allows companies to find an optimal allocation for their production considering the parameters, constraints and objectives they have defined during the model setting; on the other hand, with simulation stochastic events, such as rush orders or delays in the expected components delivery date, are taken into account, moving the production allocation analysis from a deterministic scenario to a not-deterministic one. Moreover, the comparison among simulated outputs coming from different scenarios, each one characterized by a specific set of input parameters (e.g. enabled resources, occurrence of stochastic events), can be conducted considering how a pre-defined set of Key Performance Indicators (KPIs), such as customers’ due dates compliance, advances in production and total processing cost, varies moving from a scenario to another one. Finally, the implementation of an iterative simulation-optimization framework into three different sectors (i.e. metal accessories, leather goods and footwear) has been presented, highlighting its relevance from an industrial perspective due to the fact that it represents a decision-support tool for production planners and managers that need to rapidly understand how evaluate the alignment between the gained PP&C performances and the company’s CSFs to, eventually, reallocate the already scheduled production to remain competitive in a such dynamic SC. As a future step, the information needed as input for the framework implementation could be automatically gathered through several technologies, such as Internet of Things (IoT) sensors, track and trace systems, and Radio Frequency Identification (RFId). According to this, integrating the framework implementation with third part real-data acquiring sources could allow to update in real-time the inputs and, consequently, the outputs, creating a digital twin model for the operational planning within the fashion SC.