Prioritize MRO storeroom items based on criticality and effect
ABC analysis of MRO inventory: Benefits abound for maintenance and operations analysis.
By Doug Wallace, Life Cycle Engineering
Although there’s no direct correlation between the two, an ABC analysis on MRO parts is similar to criticality rankings that reliability engineers develop for plant equipment. Criticality rankings define an asset’s relative priority in terms of its effect on plant safety, productivity, efficiency or other criteria. An ABC analysis determines the relative priority of MRO storeroom items based on their criticality to the support of operations and their potential effect on plant inventory and investment.
Completing the ABC analysis involves three key activities: prioritization, stratification and optimization. The results of the analysis provide valuable information that can be used to determine how each part should be monitored and managed from the perspectives of procurement, warehousing and strategic inventory management.
The first part of the analysis consists of rank ordering MRO storeroom parts based on a set of predefined criteria. There are several different schools of thought on the best approach to doing this. It’s important to understand the advantages and disadvantages of each before deciding which method is best suited to your particular situation.
Unlike a dependent-demand manufacturing operation with a master schedule and complete bills of material, material requirements in the MRO world are largely unpredictable. The inherent assumption in ABC analysis for MRO parts is that, at least to some degree, the past is a fairly reliable predictor of the future. Investment brokers are quick to remind us that’s not the case, but in lieu of accurate material forecast data, there’s little option. That brings up several key points to consider.
First, it’s not advisable to prioritize items by on-hand inventory value or any other variable that can change significantly from day to day, depending on issues, receipts or other transactions that take place in the storeroom. Doing so would produce dramatically different results, depending on the day you took the data snapshot.
Second, from a warehouse management perspective, there are definite benefits to analyzing inventory by activity level — number of issues or quantity of parts issued. However, using this method to prioritize inventory might negate some of the more important benefits of the ABC analysis as discussed below. If activity data is going to be used for such things as determining the most efficient stocking location for parts, the required data should be readily available from outside of the ABC process.
“Later in the life cycle, parts are likely to have historical issue data that are more heavily weighted toward the earlier part of the horizon because of declining demand over that timeframe. ”
- Doug Wallace, Life Cycle Engineering
Third, for a newly established storeroom, or where no historical data is available, it might be useful to prioritize items initially on the basis of unit cost until you develop a sufficient base of historical data.
Given these considerations, the recommended criterion for prioritization is annualized dollar value of usage. This method combines the effects of unit cost and activity level, and it provides a more stable data set obtained from a horizon of historical transaction information. Calculate the average annual usage for each item, extend it at the unit cost and then list the parts in order from highest to lowest. Several caveats apply to this approach.
First, it’s important to understand that usage is based on net quantity issued — total issues minus any returns to stock. Don’t factor in inventory adjustments, scrap or other transactions that affect the perpetual inventory balance. Not having in place proper practices that ensure timely and accurate recording of transactions when parts are issued or returned distorts the quality of the usage data and the prioritization will be degraded.
Second, it’s necessary to capture an appropriate horizon of historical data to get a reasonable representation of past usage levels. For example, in a leading-edge manufacturing environment, where product life cycles are short, a six-month horizon might be sufficient to get a fairly accurate estimate of usage.
Because in many MRO environments parts might turn over only once every few years, it’s best to use the longest possible horizon, assuming the quality of the data remain intact. Typically, a minimum of one year’s history is required, but if available, three years’ worth of data is probably sufficient.
Regardless of the time horizon, the third thing to consider is where each part is with respect to its life cycle. Parts in the early stage of their life cycles might have little historical issue data available. In this case, the calculated annualized usage likely will be understated compared to actual future requirements. On the other hand, later in the life cycle, parts are likely to have historical issue data that are more heavily weighted toward the earlier part of the horizon because of declining demand over that timeframe.