Inventory Management with Advanced Warning of Disruptions

Strategies for managing the risk of supply disruptions include inventory, dual sourcing and contingency (i.e., emergency) sourcing. Tomlin (2005) investigates the optimal mix of these strategies when a supplier faces a constant probability of a disruption. In the constant failure rate case, inventory becomes an increasingly unattractive strategy as the probability of disruption decreases (holding the supplier uptime percentage constant) because inventory needs to be carried for longer durations between disruptions. In many situations, firms receive some warning that the likelihood of a disruption at a supplier has changed. Such warnings may come in the form of feedback on the progress of labor negotiations, severe weather alerts, or the Department of Homeland Security’s terror alert level. We investigate how the presence of such warnings alters the optimal inventory policy, and therefore the optimal disruption-management strategy. We assume that the firm periodically determines its “threat level,” i.e., its estimate of the probability of a supply disruption. The firm can react by adjusting its inventory levels to provide an appropriate level of protection against an impending disruption. Previous models for inventory management under the threat of disruptions have assumed that the firm maintains a constant inventory-level target regardless of the threat level; the advantage of our approach is that inventory levels may remain low when the risk is low and increase only when the risk is heightened. In this extended abstract, we focus on a special case in which demand is deterministic, the supplier has infinite capacity, and excess inventory may be returned to the supplier at no cost. These assumptions enable closed form solutions to the inventory optimization