Design a cross-training policy to increase satisfaction and decrease cost

This research addresses a new cross-training policy to increase labors´ job satisfaction and decrease tasks´ labor cost. The cross-training plan is about how to decide which labors should be cross-trained on which tasks. A multiobjective 0-1 integer programming model is formulated for the cross-training policy. The first objective seeks to maximize average satisfaction degree (ASD), and the second objective seeks to minimize average paid salary (APS). The mathematical model is solved with particle swarm optimization algorithm (PSO). And a series of computational experiments are proceeded to analyze the factors impacting on the performance of the cross-training plan. The results indicate that with regards to ASD, the balanced preference structure is better than the extreme one, and with regards to APS, the nonuniform salary structure is better than the uniform one. Those insights will help practioners to make correct decisions.