An analysis of alternative policies to achieve schedule stability

This paper compares the effectiveness of three policies to reduce schedule instability. The policies are: (1) using safety stock at the master production schedule (MPS) level, (2) freezing a portion of the MPS, and (3) combining policies 1 and 2. The authors report simulation experiments conducted for investigating the alternative policies using single-item master scheduling in a rolling scheduling environment under uncertain demand and a variety of conditions characterized by variations in forecast error standard deviation, setup and inventory-carrying cost ratio, and lot-sizing method. The experimental results indicate that freezing a portion of the MPS produces lower lot-size costs and more stable schedules than using safety stock at the MPS level. However, freezing a portion of the MPS results in a lower service level compared to the other policies. The results also suggest that, contrary to popular belief, using safety stock may not serve to reduce schedule instability and can sometimes aggravate the problem. The authors also report additional experiments conducted to verify the sensitivity of the conclusions derived from the base experiments to the parameter values chosen for the alternative policies. The results of these sensitivity analyses indicate that the relative performance of the three policies is not affected by the particular level at which the policies are implemented.