This page is the technical reference for PlaneConnection’s predictive
maintenance system. For a step-by-step guide on using these features, see
Use Predictive Maintenance.
Alert Severity Levels
Predictive maintenance alerts use a five-level severity scale. Each level has
a defined meaning, default notification behaviour, and recommended response.
| Level | Color | Meaning | Default notification | Recommended response |
|---|
| Info | Blue | Parameter is within normal range but a notable data point was recorded. | In-app only | Note for awareness. No action required. |
| Watch | Yellow | Parameter is approaching a threshold or showing an early trend deviation. | In-app, email | Monitor at increased frequency. Review at next scheduled check. |
| Caution | Orange | Parameter has exceeded a Watch threshold or trend slope is statistically significant. | Email, push | Schedule an inspection within the next scheduled maintenance window. |
| Warning | Red | Parameter has exceeded a Caution threshold. Immediate maintenance attention warranted. | Email, SMS, push | Ground aircraft pending maintenance review unless cleared by DOM. |
| Critical | Dark red | Parameter has exceeded a Warning threshold or a safety-critical exceedance occurred. | Email, SMS, push | Ground aircraft immediately. Notify DOM and PIC. |
Alert severity levels are advisory. Airworthiness decisions must be made by a qualified aviation
maintenance technician (AMT) or Director of Maintenance in accordance with the aircraft’s approved
maintenance program and applicable FARs.
Monitored Parameters
Turbine Engine Parameters
The following parameters are monitored for turbine (jet and turboprop) engines.
Parameter names follow SAE ARP4102 conventions where applicable.
| Parameter | Abbreviation | Unit | Description |
|---|
| Fan speed (N1) | N1 | % RPM | Low-pressure compressor/fan rotational speed. Primary power setting. |
| Core speed (N2) | N2 | % RPM | High-pressure compressor rotational speed. |
| Exhaust gas temperature | EGT | °C | Temperature at turbine exit. Primary thermal health indicator. |
| Inter-turbine temperature | ITT | °C | Temperature between turbine stages (turboprop engines). |
| Fuel flow | FF | lbs/hr or kg/hr | Fuel consumption rate. Correlates with power and efficiency. |
| Oil pressure | OIL PRESS | PSI | Engine oil system pressure. |
| Oil temperature | OIL TEMP | °C | Engine oil temperature. Elevated values indicate thermal stress. |
| Oil quantity | OIL QTY | quarts or litres | Oil consumption between services. Rising consumption indicates wear. |
| Vibration (N1/N2) | VIB | inches/sec or EU | Engine vibration level. Elevated levels indicate imbalance or bearing wear. |
| Torque (turboprop) | TRQ | ft·lb or % | Shaft output torque (turboprop engines). |
| Bleed air pressure | BL AIR | PSI | Compressor bleed air pressure for environmental and pneumatic systems. |
Piston Engine Parameters
| Parameter | Abbreviation | Unit | Description |
|---|
| Cylinder head temperature | CHT | °C | Temperature of cylinder heads. Key detonation and cooling indicator. |
| Exhaust gas temperature | EGT | °C | Exhaust temperature. Used for mixture management and trend analysis. |
| Oil pressure | OIL PRESS | PSI | Engine oil system pressure. |
| Oil temperature | OIL TEMP | °C | Engine oil temperature. |
| Manifold pressure | MAP | in. Hg | Induction manifold pressure. Power setting indicator. |
| RPM | RPM | RPM | Engine rotational speed. |
| Fuel flow | FF | GPH | Fuel consumption rate. |
Airframe Parameters
| Parameter | Description |
|---|
| Landing gear cycles | Cumulative extension/retraction cycles tracked for life-limited components. |
| Pressurisation cycles | Cumulative pressurisation cycles for fatigue tracking. |
| Flight hours | Total airframe hours since new and since last major inspection. |
| Fuel imbalance | Left/right tank imbalance exceeding configured threshold. |
| Hydraulic pressure | System hydraulic pressure on equipped aircraft. |
Trend Analysis Methodology
PlaneConnection’s trend analysis uses two complementary statistical methods to
detect parameter degradation. Both methods run on each parameter independently.
Linear Regression
Linear regression fits a straight line through the most recent N data points
(configurable window, default: 30 flight legs) and computes the slope. A
positive slope on a rising-bad parameter (e.g., EGT) indicates a degrading
trend even if no individual reading has exceeded a threshold.
| Slope magnitude (EGT example) | Classification | Recommended action |
|---|
| < 0.5 °C/flight hour | Stable | Normal monitoring |
| 0.5–1.5 °C/flight hour | Watch | Increase monitoring cadence |
| 1.5–3.0 °C/flight hour | Caution | Schedule inspection |
| > 3.0 °C/flight hour | Warning | Immediate review |
Exponential Smoothing
Exponential smoothing applies a weighted average where recent readings have
more influence than older readings. The smoothing factor α (alpha) controls
the decay rate: higher α means faster response to recent changes.
PlaneConnection uses a default α of 0.2 for most parameters, meaning
approximately the last 5 readings contribute meaningfully to the smoothed value.
For highly volatile parameters (vibration), α defaults to 0.3.
The smoothed value is compared against threshold bands to determine alert level.
Exponential smoothing is less sensitive to single-point spikes than raw
thresholds but more responsive to sustained trends than long-window linear
regression.
Combined Alert Generation
An alert is generated when either the regression slope or the smoothed value
indicates a threshold exceedance. The higher-severity result takes precedence.
This dual-method approach reduces both false positives (single outlier spikes)
and false negatives (slow degradation that stays below instantaneous thresholds).
Threshold Configuration Options
Each monitored parameter on each aircraft can be configured independently.
Threshold settings are found under **Ops > Predictive Maintenance > [Aircraft]
Configure Thresholds**.
| Setting | Type | Default | Description |
|---|
| Alert direction | Rising/Falling/Both | Parameter-specific | Whether the alert fires when values rise, fall, or deviate in either direction. |
| Watch threshold | Numeric | 90% of Caution | Value at which a Watch alert is generated. |
| Caution threshold | Numeric | Per aircraft type | Value at which a Caution alert is generated. |
| Warning threshold | Numeric | Per aircraft type | Value at which a Warning alert is generated. |
| Critical threshold | Numeric | Per aircraft type | Value at which a Critical alert is generated. |
| Regression window | Integer (legs) | 30 | Number of flight legs used for slope calculation. |
| Smoothing factor (α) | 0.0–1.0 | 0.2–0.3 | Exponential smoothing decay rate. Higher = more responsive to recent data. |
| Consecutive exceedances | Integer | 1 | Number of consecutive readings above threshold before alerting. Reduces noise. |
| Baseline source | Enum | Aircraft history | Comparison baseline: aircraft-specific history, fleet average, or manufacturer limits. |
| Suppress below n legs | Integer | 10 | Do not alert until this many legs have been recorded (avoids false alerts on new aircraft). |
Start with the default thresholds and tighten them over time as you accumulate baseline data for
each aircraft. The default thresholds are conservative and may generate Watch-level alerts on
healthy aircraft during the first few months of use.
Data Sources
Predictive maintenance parameters are populated from the following data sources:
| Source | Parameters populated | Update frequency |
|---|
| ACARS / digital OOOI data | Flight legs, departure/arrival times | Per flight event |
| EFB data import (CSV/JSON) | Engine parameters per leg (N1, N2, EGT, FF, oil) | Manual or automated upload |
| FOQA data import | Full parameter set at recording frequency | Per flight (after download) |
| Manual entry (maintenance log) | Oil quantity, maintenance actions, component replacements | Per maintenance event |
| Sensor integrations (API) | Real-time or near-real-time parameter streams | Per configured interval |
FOQA data integration requires the Use FOQA feature to be enabled and data to be uploaded per
the Use FOQA guide. Manual parameter entry is always available as a
fallback for operators without automated data sources. Use Predictive Maintenance
Step-by-step guide for using the predictive maintenance dashboard.
Use Fleet Health
Fleet-wide health overview combining predictive alerts and due items.
Use FOQA
Flight Operational Quality Assurance data that feeds engine trend monitoring.
Track Due Items
Schedule-based maintenance tracking alongside condition-based predictive alerts.
