Parts cannibalization is one of the most operationally normalized practices in MRO that almost no planning system accurately captures. When a planner authorizes the removal of a serviceable unit from a grounded or low-priority aircraft to resolve a shortage on a revenue-critical aircraft, the immediate problem is solved. The part is in service. The flight departs. The AOG log never shows an event.
What the log also does not show is the demand that just occurred. The removal happened. The part was consumed from an aircraft's serviceable installation, not from inventory. And the aircraft it was taken from is now missing a component that will need to be replaced before it can fly its next assignment.
Why Cannibalization Disappears from Demand Records
The mechanics of how cannibalization escapes demand tracking vary by MRO system, but the core problem is consistent: most systems record parts in terms of inventory transactions, not as operational demand events. When a part comes off the shelf and gets installed on an aircraft, the inventory decreases and a demand event is logged. When a part is removed from one aircraft and installed on another, the inventory balance does not change. The part was never on the shelf. The demand tracking logic never fires.
In AMOS and CESIUM, cannibalization removals are typically coded as aircraft-to-aircraft transfers rather than consumptions. They may appear in work order records, but they do not flow into the demand history that feeds reorder calculations. The planner looking at their reorder report sees no demand signal for the component that just moved between aircraft. The model does not know the shortage occurred.
This creates a systematic undercount of actual demand. If a component is cannibalized once per quarter across a 20-aircraft fleet, the model is missing four demand events per year. For a slow-moving component with an annual demand rate of eight to twelve units, missing four events is a 35 to 50 percent undercount of true demand. The reorder point that comes out of that undercount will consistently leave the fleet short.
The Secondary Cost: the Donor Aircraft
The aircraft that donated the part is now in a deferred defect state. In most regulatory frameworks, an aircraft cannot be dispatched with a known unresolved maintenance item unless it is covered by a minimum equipment list (MEL) deferral. If the cannibalized component is MEL-eligible, the donor aircraft can continue flying. If it is not, the donor aircraft becomes the new AOG.
This is the substitution that the AOG log misses. The original shortage resolved on aircraft A. Aircraft B is now on the ground waiting for the component to be sourced from a vendor or returned from the repair cycle. The net AOG exposure for the fleet did not decrease; it transferred. The log shows one AOG closed without showing the one that opened.
Fleet-level AOG risk analysis that relies on AOG log counts rather than actual parts availability is therefore systematically optimistic. The log is measuring when planners successfully managed AOG risk on individual aircraft, not whether fleet-level parts availability was adequate.
How Cannibalization Inflates MTBR Calculations
There is a secondary planning distortion that flows from uncaptured cannibalization events, which we touched on in our earlier article on MTBR limitations. When a part is removed from an aircraft via cannibalization, that removal event may still appear in the work order history even if it does not appear in demand tracking. If the MTBR calculation is based on total removals from work order records rather than on demand-driven removals, cannibalization events inflate the apparent removal rate.
A component that genuinely fails once every 900 flight hours but is cannibalized twice per year in a 15-aircraft fleet will show an apparent removal rate in the work order data that is higher than its true failure rate. The MTBR comes out lower than the failure pattern warrants. The planner calibrates safety stock against this inflated rate. The result is more inventory than the actual failure rate requires, but not necessarily inventory available at the right aircraft at the right time, because the shortage pattern is driven by distribution and timing, not by aggregate volume.
Identifying Cannibalization Events in Existing Records
Most MRO systems do contain information that allows retrospective identification of cannibalization events, even when they are not explicitly flagged. Aircraft-to-aircraft transfer codes in work order records, non-routine work orders where no shelf transaction accompanies the installation record, and cases where an aircraft shows a component removal immediately followed by another aircraft showing an installation without a corresponding inventory decrement are all identifiable patterns.
When we build a demand model from a new operator's records, scanning for these patterns is one of our first data quality steps. The goal is not to reconstruct a perfectly clean demand history from the start, but to understand the volume of cannibalization-driven demand that the standard demand records have missed. That baseline tells us how much the planning model needs to adjust its demand estimates upward for the affected components.
For a 12-aircraft Embraer E175 operator we worked with in late 2024, the analysis found that approximately 18% of avionics LRU removals in their two-year history were cannibalization events that had not been captured as demand events. For the three most-cannibalized part numbers, the true annual demand rate was 25 to 40% higher than the demand records indicated. Their reorder points were calibrated to the wrong base rate.
What Reducing Cannibalization Actually Requires
We should be direct about the organizational dimension here. Cannibalization is often a symptom of chronically inadequate stock levels rather than a habit that can be eliminated by process changes alone. If a planner does not have enough units on hand to fill demand from shelf inventory, cannibalization is a rational response. Telling them to stop cannibalizing without addressing the underlying stock gap will result in delayed flights instead of cannibalized parts.
The sequence matters. The first step is accurate demand capture: identify what the true demand rate is, including cannibalization-adjusted counts. The second step is calibrating stock levels to that true demand rate. The third step is building the monitoring that detects when cannibalization is occurring in near real-time rather than in a retrospective quarterly audit. With adequate stock levels maintained against accurate demand, cannibalization frequency tends to decrease organically because the shortage events that drive it become less frequent.
Cannibalization does not disappear from an operation because a planner decides to stop doing it. It decreases when the parts are there when the aircraft needs them.
See Aerotrax on your own fleet data
30-day pilot on your removal history. CSV upload, no integration project required.
Request a pilot