Landing gear systems generate some of the most predictable demand patterns in MRO parts planning, yet they remain among the most common contributors to AOG delays at regional carriers. The reason is not a lack of data. It is that the signals in that data are distributed across multiple record types and require active interpretation rather than passive reporting.
What follows is a description of five removal data signals we have found consistently predictive when extracted and structured from landing gear maintenance records. These are patterns experienced planners know intuitively. The value in naming them explicitly is in building demand models that can detect them without relying on a single planner's memory of the last time the pattern appeared.
Signal 1: Landing Cycle Accumulation Rate by Tail Number
Landing gear components degrade primarily as a function of landing cycles, not flight hours. A regional aircraft doing 8 to 10 cycles per day is accumulating stress on its gear at three times the rate of a long-haul aircraft doing 2 to 3 cycles per day, even if both are flying similar block hours.
The predictive signal here is not just total accumulated cycles, but the current rate of cycle accumulation relative to the component's historical removal interval. If a nose gear steering actuator on a specific tail number has been averaging 2,400 landing cycles between removals and that tail number is approaching 2,100 cycles since last removal, the probability of removal in the next 60 days is substantially higher than the fleet average would suggest.
Cycle accumulation data is almost always available in the MRO system, but it is often not linked to the demand forecast in any active way. The planner has to manually check. A model that reads cycle data and raises a removal probability flag does this automatically, without requiring the planner to remember which tail numbers to monitor.
Signal 2: Prior Defect Log Clustering
Unscheduled landing gear removals are rarely isolated events. In most landing gear system architectures, a component approaching failure begins to generate defect log (DEL) entries in the preceding weeks or months. Hydraulic seepage notes, extension-retraction time anomalies, and ground sensor discrepancy write-ups are common precursor entries before an actuator or control valve removal.
The signal here is the rate and category of DEL entries associated with a specific gear system component on a specific tail number over the preceding 45 to 90 days. An aircraft accumulating two or more related DEL entries in a 30-day window is statistically more likely to generate an unscheduled removal than one with no recent entries, even if both aircraft have identical cycle counts.
DEL data is rich but underused in planning because it lives in a different data table from the removal records. Connecting them requires a data join that most planning systems do not perform automatically.
Signal 3: Brake Assembly Removal Clustering as a System Indicator
When brake assemblies are removed at an elevated rate across multiple aircraft in a fleet, it is sometimes a signal for demand in adjacent gear system components. Elevated brake temperatures, hard landing events, and abnormal brake wear often co-occur with increased hydraulic system stress. A gear system running under elevated hydraulic demand due to brake compensator activity shows higher seal and actuator removal rates in the 60 to 90 days following the brake event cluster.
This is a system-level signal, not a component-level one. Identifying it requires looking at removal events across component families rather than tracking each component in isolation. A planner watching brake removals as a stand-alone metric will not see this. A model that monitors cross-component co-occurrence patterns can.
Signal 4: Runway Surface Composition at High-Frequency Stations
Aircraft operating predominantly from runways with rough or contaminated surface conditions show higher gear component removal rates than the same aircraft type operating from well-maintained major hub runways. This is not a dramatic effect in the first two years of service, but it accumulates meaningfully over five to eight years of concentrated operation at specific station types.
For operators with stable route networks, the high-frequency station composition is a useful segmentation variable when building removal-rate estimates by tail number. An aircraft spending 70% of its cycles at a regional airport with known pavement quality issues should carry a different base removal rate for gear bushings and shimmy damper components than a similar aircraft flying primarily between major hubs.
This variable is not available in MRO records directly. It requires cross-referencing route data with the removal history, which is an additional data join. For operators willing to make that connection, it adds forecasting accuracy for gear-intensive components.
Signal 5: Overhaul Shop Return-to-Service Rate Variance
Landing gear overhauls are typically major shop visits with lead times measured in months. When a gear set comes back from overhaul, the components installed during overhaul carry a reset clock. But the quality of that reset varies between overhaul facilities and between overhaul visits, depending on which components were renewed versus inspected-and-continued.
The signal here is the variance in time-to-first-unscheduled-removal after different overhaul events. A gear set that returns from overhaul and generates its first removal event within 400 cycles is showing a different post-overhaul reliability profile than one that runs 2,000 cycles before requiring any intervention. Tracking this variance by overhaul facility and overhaul type reveals whether certain overhaul quality patterns predict elevated near-term demand after shop return.
This is a more sophisticated data requirement than the previous signals. It requires clean linkage between overhaul completion records and subsequent removal events by tail number, with overhaul facility as a variable. Not every operator has this data in a form that supports easy analysis, but for those with established heavy maintenance programs, it is one of the more powerful demand predictors available.
Connecting the Signals to Planning Action
The goal of identifying these signals is not pattern recognition for its own sake. It is generating a 30-to-60-day demand probability for specific gear components by tail number, so that reorder decisions are made before the stockout, not after it.
A regional carrier operating 16 aircraft at a hub in the Mountain West ran a Aerotrax pilot against two years of their landing gear removal history in early 2025. Connecting cycle accumulation data, DEL clustering, and post-overhaul removal variance produced a 90-day gear component demand forecast that identified three tail numbers as high-probability removal candidates within the quarter. Two of the three generated removals within the forecast window. The third did not remove in the quarter but entered a heavy check where the relevant component was renewed on schedule.
We should be clear about what these signals do not do: they do not eliminate all surprise removals. Gear system failures can result from foreign object damage, hard landing events, and manufacturing anomalies that leave no precursor trace in the data. The signals described here address the predictable portion of the demand curve. For some component categories that portion is 60 to 70 percent of annual demand volume. Getting that portion right substantially reduces the AOG exposure from gear-related inventory gaps.
See Aerotrax on your own fleet data
30-day pilot on your removal history. CSV upload, no integration project required.
Request a pilot