Designing Resilient Factory Layouts Through Data-Driven Scenario Analysis with an Industrial Application

Developing resilient and robust facility layouts is a long-term strategic problem for contemporary manufacturing systems, particularly in settings with operational uncertainty and spatial complexity. When departments are assigned to multiple layout blocks of different area requirements and material flows are prone to disruption, conventional deterministic models are not able to provide robust solutions. This paper bridges this gap by introducing a scenario-based robust facility layout model as a mixed-integer linear programming (MILP) problem. We introduce an MILP formulation that couples the unequal-area facility layout problem with a capacitated network flow problem. Unlike standard models that assume direct paths, this approach explicitly models interblock logistics through capacity-constrained gateways (ramps). The model employs a penalty-based robust objective function that integrates expected cost minimization with stability deviations and soft capacity constraints, allowing the identification of feasible solutions even under compound disruption scenarios where rigid models would fail. The proposed approach is demonstrated on a real-life application case from a leading European home appliance manufacturer based in Turkey, where the plant layout must fit changing operational conditions and capacity restrictions. Leveraging actual production data and expert knowledge such as observations of material surges and production shutdowns, realistic disruption scenarios are defined for the model to develop solutions that remain feasible and efficient in different operating states. Simulation-based verification shows that optimized layouts preserve robust performance under dynamic and uncertain flow conditions. The proposed layout reduced total material handling distance by 19.3% and increased production area utilization by 31% compared with the current layout. The study provides a scalable and data-driven approach to industrial facility planning and demonstrates the practical applicability of robust optimization to increase resilience under uncertainty in production settings.

History: This paper was refereed.

Funding: This work was supported by the Scientific and Technological Research Council of Türkiye [Grants 118C128 and BIDEB 2244].