Due to the nature of the manufacturing and support activities associated with long life cycle products, the parts that products required need to be dependably and consistently available. However, the parts that comprise long lifetime products are susceptible to a variety of supply chain disruptions. In order to minimize the impact of these unavoidable disruptions to production, manufacturers can implement proactive mitigation strategies. Two mitigation strategies in particular have been proven to decrease the penalty costs associated with disruptions: second sourcing and buffering. Second sourcing involves selecting two distinct suppliers from which to purchase parts over the life of the part’s use within a product or organization. Second sourcing reduces the probability of part unavailability (and its associated penalties), but at the expense of qualification and support costs for multiple suppliers. An alternative disruption mitigation strategy is buffering (also referred to as hoarding). Buffering involves stocking enough parts in inventory to satisfy the forecasted part demand (for both manufacturing and maintenance requirements) for a fixed future time period so as to offset the impact of disruptions. Careful selection of the mitigation strategy (second sourcing, buffering, or a combination of the two) is key, as it can dramatically impact a part’s total cost of ownership.
This paper studies the effectiveness of traditional analytical models compared to a simulation-based approach for the selection of an optimal disruption mitigation strategy. A verification case study was performed to check the accuracy and applicability of the simulation-based model. The case study results show that the simulation model is capable of replicating results from operations research models, and overcomes significant scenario restrictions that limit the usefulness of analytical models as decision-making tools. Four assumptions, in particular, severely limit the realism of most analytical models but do not constrain the simulation-based model. These limiting assumptions are: 1) no fixed costs associated with part orders, 2) infinite-horizon, 3) perfectly reliable backup supplier, and 4) disruptions lasting full ordering periods (as opposed to fractional periods).
Doctors’ Extended Shifts as Risk to Practitioner and Patient: South Africa as a Case Study