Jointly optimizing assembly line feeding and supermarket layouts
Description
This dataset supports a study on optimizing the assembly line feeding problem (ALFP) in mixed-model assembly systems. The underlying hypothesis is that jointly optimizing feeding policy selection, cell assignment (regular vs. line-integrated), cell placement, and cell sizing leads to lower overall costs compared to treating cell placement and sizing as fixed parameters. The input data consist of 36 artificially generated problem instances, based on established literature, representing assembly systems with seven stations, 49–74 parts, and 19–35 part families. Parameters include delivery frequencies, processing requirements, distances, and cost components related to replenishment, preparation, transportation, and usage. The data are structured to reflect realistic intralogistics conditions and enable evaluation of different feeding policies such as line stocking, boxed supply, sequencing, and kitting. The output data provide the optimized system configuration for each instance, including the selected feeding policy per part, assignment to regular or line-integrated cells, as well as cell locations and sizes. Results show that explicitly incorporating cell placement and sizing decisions can significantly reduce total system costs, particularly by lowering transportation and space-related expenses. The dataset can be used to interpret trade-offs between different logistics design choices, reproduce the study’s findings, and benchmark alternative optimization models for assembly line feeding and intralogistics planning. The Input folder contains the problem instances. Within the Output folder, the Output-Gurobi folder presents the model results without any enhancements, and the Output-VISBC folder provides the results obtained after incorporating enhancements such as valid inequalities and symmetry-breaking constraints. Note that the results in the Output-VISBC folder are used for comparison in Sections 5.4 and 5.5. Specifically, the Output-VISBC folder corresponds to the cell-based (CB) approach in Table 5 and the base case in Table 7. The Output-Heuristic folder contains the results for the practice-oriented decision rule (DR) presented in Table 5. Finally, the Output-Case1 to Output-Case5 folders present the results for the five cases discussed in Tables 6 and 7 in Section 5.5.
Files
Institutions
- Ghent UniversityFlanders, Ghent