A Decision Support Framework for Manufacturing Improvement and Relocation Prevention in Thailand: Supply Chain Perspective

The low economic growth and competition among neighbouring countries has caused Foreign Direct nvestments (FDIs) to relocate their businesses. In order to prevent further business relocation, this aper proposes an integrated framework based on the supply chain to help analyse decision making for plant situations and enhance manufacturing perrformance. The context of this perspective is pplied to manufacturers located in the industrial state region of Lumphun province, Thailand. Data collection and review of literature was used to identify the factors that influence industrial investment. The SCOR model was used to define the parameters, which here then used in Arena simulation. The simulation needs to describe factors affected on industrial erformance. From this result, an integrated analysis model was built and the importance of supply chain collaboration was identified. A Multi Agent System (MAS) was suggested to enhance of the effective nteraction between supply chain’ agents. It is in order to mitigate risks among them.

💡 Research Summary

The paper addresses the growing problem of foreign direct investment (FDI) relocations in Thailand, driven by sluggish economic growth and intensified competition from neighboring countries. To counteract further plant closures and to boost manufacturing performance, the authors develop an integrated decision‑support framework that is rooted in supply‑chain theory and operational simulation.

First, a comprehensive literature review and field data collection (surveys and interviews) were conducted in the industrial region of Lampang province. This effort identified the key determinants of industrial investment decisions, including government policy, infrastructure quality, labor costs, technological capability, and, most critically, the degree of supply‑chain collaboration among firms, logistics providers, and public agencies.

The authors then mapped these determinants onto the Supply Chain Operations Reference (SCOR) model, which structures supply‑chain activities into five core processes: Plan, Source, Make, Deliver, and Return. For each process, quantitative performance metrics (e.g., OTIF, lead time, cost per unit, inventory turns) and qualitative factors (e.g., trust, information sharing) were defined.

Using the Arena discrete‑event simulation platform, the research team built a series of “what‑if” scenarios that varied the identified factors. The simulation demonstrated that low collaboration levels create a “bubble effect,” where minor demand fluctuations or supply disruptions rapidly cascade into severe system inefficiencies, higher costs, and longer lead times. Conversely, enhanced collaboration mechanisms—joint inventory management, shared forecasting, and contract‑based negotiations—significantly improve system resilience, reduce total cost of ownership, and increase service levels.

Building on these results, the paper proposes an integrated analysis model that combines (1) external environment variables (policy, infrastructure, labor), (2) internal SCOR‑based performance indicators, and (3) a collaboration index. The model employs a hybrid approach: multivariate regression captures static relationships, while the simulation outputs feed dynamic adjustments. Crucially, the authors embed this model within a Multi‑Agent System (MAS). Each supply‑chain participant (supplier, manufacturer, logistics provider, government) is represented as an autonomous agent with its own objectives, yet governed by interaction rules that enforce information exchange and coordinated decision‑making. The MAS continuously ingests real‑time data, updates simulation parameters, and generates scenario‑based recommendations for risk mitigation and investment retention.

Key insights emerging from the study are:

1. Supply‑chain collaboration infrastructure carries the highest weight among investment‑retention factors, surpassing traditional considerations such as tax incentives or labor cost differentials.
2. Single‑firm decision frameworks are insufficient in a highly interconnected environment; a multi‑stakeholder, collaborative decision‑making paradigm is essential for robustness.
3. MAS‑enabled simulation provides a rapid “predict‑adjust‑respond” cycle, allowing policymakers and managers to anticipate the ripple effects of policy changes, infrastructure upgrades, or market shocks before they materialize.
4. While the case study focuses on Lampang, the methodology is scalable and can be adapted to other Southeast Asian manufacturing hubs facing similar relocation pressures.

In conclusion, the paper delivers a novel, supply‑chain‑centric decision‑support architecture that merges SCOR‑based performance modeling, Arena simulation, and MAS coordination. This framework equips Thai manufacturers and government officials with actionable intelligence to prevent plant relocations, improve operational efficiency, and sustain competitive advantage in a volatile regional landscape. Future research directions include integrating live IoT data streams into the MAS for real‑time optimization and extending the policy simulation component to evaluate large‑scale economic development strategies.