Worker flexibility in a parallel dual resource constrained job shop

In this paper we investigate cross-training policies in a dual resource constraint (DRC) parallel job shop where new part types are frequently introduced into the system. Each new part type introduction induces the need for workers to go through a learning curve. A cross-training policy relates to the level of multi-functionality, the pattern of skill overlaps, and the distribution of skills among workers. Our results show that the frequency of new part type introduction needs to be considered when selecting a cross-training policy. The frequency of new part type introduction has impact on the optimal level of cross-training, the relative advantage of skill overlaps, and the extent to which a cross-training policy should incorporate differences between workers. These findings are important for many DRC job shops, especially for manufacturing cells in which all changes, associated with a part type introduction, are concentrated in one cell.