Replication package for "Robust Machine Prioritization in Semiconductor Manufacturing Using LOPCOW–RAM: Cross-Method and Monte Carlo Robustness Analysis"

Replication package for the article “Robust Machine Prioritization in Semiconductor Manufacturing Using LOPCOW–RAM: Cross-Method and Monte Carlo Robustness Analysis” (submitted to Applied Operations and Analytics). The package contains: (1) R code for a 180-day reduced-order discrete-event simulation of the six tools of toolgroup Dielectric_BE_21 in the public SMT2020 Dataset 2 (low-volume/high-mix) testbed — a V7 calibration run and a V8 closed-loop run in which arriving lots are assigned to machines by a state-dependent CRITIC–TODIM rule; (2) the simulation output workbooks, including event logs, snapshots, correlation diagnostics and the representative 6 × 13 decision matrix (mean over 1,100 eligible decision epochs within ±7 days of the 75% horizon point); (3) the decision matrix as CSV; and (4) an Excel workbook with all MCDM calculations: LOPCOW weighting, RAM ranking, 13 benchmark methods (TOPSIS, VIKOR, EDAS, CoCoSo, CRADIS, CODAS, COBRA, MABAC, MARCOS, WASPAS, ARAS, MOORA, PROMETHEE II), six alternative weighting schemes, leave-one-criterion-out, 78 one-at-a-time weight changes, 10,000 Monte Carlo perturbations of the weights and of the decision matrix, rank-reversal tests, the RAM shortfall decomposition, and agreement statistics (Spearman, WS, Kendall’s W, Friedman–Nemenyi). The original SMT2020 workbook is not redistributed; see README.md and 01_input/ for download instructions and a SHA-256 checksum. Workflow, folder structure and a map of the Excel workbook are documented in README.md.