Multi-project scheduling with multi-skilled workforce assignment considering uncertainty and learning effect for large-scale equipment manufacturer

Abstract

This study addresses a resource-constrained multi-project scheduling and multi-skilled workforce assignment problem for the large-scale equipment manufacturing industry. The goal is to help the production manager perform effective planning and decrease the production cost under delivery constraints. The deterministic duration is used in most past investigations in resource-constrained multi-project scheduling problems. Uncertainty in project execution is neglected. Additional related factors are included in the current study to be close to practical circumstances. The required processing duration and material arrival time are modeled as stochastic variables. The heterogeneous skill efficiency of the internal workforce is considered with the learning effect. Skill efficiency grows with the accumulation of experience. Actual processing time is calculated on the basis of skill efficiency and stochastic process duration. The external workforce is hired when the internal workforce fails to satisfy processing demand. The objective is to minimize the expected integrated cost, which is the sum of tardiness penalty and external workforce cost. A genetic algorithm with a heuristic workforce assignment method is also developed, and a case study is illustrated to explain the scheduling result. The proposed approach with uniform crossover and local search mechanism outperforms other comparative methods by approximately 10.34% considering total costs. Furthermore, the Taguchi method of design of the experiment is conducted to find the optimal parameter setting efficiently. Afterward, the sensitivity analysis of production-related parameters is discussed.