Human resource planning in knowledge-intensive operations: A model for learning with stochastic turnover

Because of rapid evolution in product and process technology, many operations in the manufacturing and service industries in recent years require workers to acquire and maintain more extensive ‘knowledge stock’ than before. In this paper, we address the human resource planning in these knowledge-intensive operations. We focus on the management of knowledge mix, that is, the mix of workers in different knowledge levels. This research was conducted in a semiconductor equipment manufacturing plant that uses an assembly line to achieve high productivity. However, the plant also needs to increase flexibility to deal with a high degree of product customization, frequent technology and product changes, and relatively low volume of production, which collectively require a high level of knowledge development for each worker. We extend methodology developed for managing production and work-in-process inventory levels for a manufacturing system that is subject to random production yields. However, the structure of our problem is different, and thus requires a separate mathematical development. Our results indicate that the company we studied underestimated the ideal number of workers in the higher knowledge levels in the steady state. But this problem, by itself, can be taken care of by a good intuitive heuristic. We then demonstrate that our control rule is superior to the good intuitive rule from the point of view of additional stability that is obtained from less variability in the work force levels. We offer managerial implications of this additional stability using our computational results.