Adapting lot-sizing techniques to stochastic demand through production scheduling policy

Production-inventory systems that experience significant demand variations from period to period often use deterministic lot-sizing techniques to schedule production runs. If demand is uncertain, safety stock is included to buffer against stockouts. In this paper we propose a 'dual-buffer' production scheduling policy and show that it has superior performance characteristics to the single buffer policies currently in use. The second buffer is a simple calculation based on prior work in stochastic inventory theory. Methodologically, data envelopment analysis (DEA) is used to analyze results. This represents an unusual use of DEA