Stochastic skill-based manpower allocation in a cellular manufacturing system

• In this paper, stochastic skill-based manpower allocation problem is addressed, where operation times and customer demand are uncertain.
• A four-phased hierarchical methodology that consists of three stochastic nonlinear mathematical models and Monte Carlo simulation is developed.
• Each worker's expected and standard deviation of processing time on each operation is considered individually.
• The hierarchical approach allows decision makers to perform manpower level determination, cell loading and individual worker assignment with respect to the desired risk level.

In this paper, stochastic skill-based manpower allocation problem is addressed, where operation times and customer demand are uncertain. A four-phased hierarchical methodology is developed. Egilmez and Süer's [1] stochastic general manpower allocation problem is extended such that each worker's individual performance is considered for a more accurate manpower allocation to manufacturing cells to maximize the production rate. The proposed methodology optimized the manpower levels, product-cell formations and individual worker assignment hierarchically with respect to a specified risk level. Three stochastic nonlinear mathematical models were developed to deal with manpower level determination, cell loading and individual worker assignment phases. In all models, processing times and demand were assumed to be normally distributed. Firstly, alternative configurations were generated. Secondly, IID sampling and statistical analysis were utilized to convert probabilistic demand into probabilistic capacity requirements. Thirdly, stochastic manpower allocation was performed and products were loaded to cells. In the final phase, individual worker assignments were performed. The proposed methodology was illustrated with an example problem drawn from a real manufacturing company. The hierarchical approach allows decision makers to perform manpower level determination, cell loading and individual worker assignment with respect to the desired risk level. The main contribution of this approach is that each worker's expected and standard deviation of processing time on each operation is considered individually to optimize the manpower assignment to cells and maximize the manufacturing system production rate within a hierarchical robust optimization approach.