Balancing Flexibility and Inventory in Workforce Planning with Learning
When planning for production, it is important to recognize the dynamic and ever changing nature of a workforce. It is commonly accepted that individuals tend to learn through experience and task repetition which in turn results in higher productivity. In this paper, we consider a model of assigning tasks to individuals in a serial production setting and examine the impact on revenue and lost sales of several different factors. We create a rolling horizon experimental setup in which we embed a one-step lookahead. The one-step lookahead requires the solution of a stochastic program for which we introduce a matheuristic approach.
Our analysis evaluates the relative significance of modeling learning, heterogeneity, constant work in progress, and uncertainty in demand. Further, seeking to identify relationships between individual differences and levels of specialization and cross training, we asses the final accumulated experience of workers. Additionally, we measure levels of inventory, its importance in the different demand variation realizations, and at different points in the planning horizon.