Data-Driven Cash Replenishment Planning of Recycling ATMs with Out-of-Cash and Full-of-Cash Risks

Abstract

When banks replenish the cash held in automated teller machines (ATMs) it is crucial for them to reduce operational costs while maintaining service level. This article studies the replenishment planning for recycling ATMs, which allow cash deposits to be made as well as withdrawals. The problem is formulated as a special (s, S) inventory model with two safety stocks corresponding to out-of-stock and full-of-stock risks, based on which the ATMs to be replenished each day and the replenishment amount are determined. Experiments with real data show that the model can significantly reduce costs and improve the overall service level.