Benchmark-derived data for designing cross-training in recoverable hybrid flowshop schedules under limited worker reassignment

Description

This dataset provides model-ready inputs for studying recoverable hybrid flowshop scheduling with cross-training and limited worker reassignment. It combines 80 openly available permutation-flowshop benchmark instances, drawn equally from the Taillard and VRF families, with deterministic workforce and disruption extensions. Each benchmark matrix is represented at three initial cross-skill densities, yielding 240 workforce-aware configurations. The original job-stage processing times are preserved without rescaling, perturbation, or imputation. Added fields describe parallel staffed positions, incumbent and reserve workers, worker-stage efficiencies, initial qualifications, candidate training actions, integer training costs, and in-sample and out-of-sample absence scenarios. All workforce attributes and scenarios are controlled synthetic data generated with master seed 24051993; they do not represent observed employees or an identifiable organization. The release contains seven relational CSV tables, a data-only Excel workbook, a variable-level data dictionary, source and license information, and SHA-256 integrity manifests. Optimization outcomes, selected training plans, objective values, runtimes, figures, manuscript files, and analysis code are deliberately excluded.

Steps to reproduce

The CSV files constitute the canonical release. Link the seven data tables using instance_id and link scenarios to their absent workers using the composite key (instance_id, scenario_id). Worker-specific processing time is obtained by dividing base_processing_time by efficiency. The original benchmark processing times are unchanged. The workforce and disruption layers were generated deterministically with master seed 24051993. Variable definitions and source provenance are provided in data_dictionary.csv and source_manifest.csv.

Institutions

• Istanbul Health and Technology University

  Istanbul, Istanbul

Categories

Licence