Glossary
This glossary defines technical terms used throughout the LogHat documentation and in the flight analysis reports. Definitions are written to be accessible to non-technical users while remaining accurate enough to be useful to engineers.
ArduPilot
ArduPilot is an open-source flight control software project that runs on a wide range of flight controller hardware (Pixhawk, Cube, Matek, and others). It is one of the most widely adopted autopilot firmware stacks in the commercial and research drone industry, supporting multirotors, fixed-wing aircraft, rovers, boats, and underwater vehicles.
ArduPilot includes a system called DataFlash that records detailed binary log files (.bin) to a connected SD card during flight. These log files contain the complete record of sensor data, controller outputs, and system messages from the flight, and are the primary input format accepted by LogHat.
Variants of ArduPilot include: ArduCopter (multirotors), ArduPlane (fixed-wing), ArduRover (ground vehicles), and ArduSub (underwater vehicles).
PX4
PX4 is another widely used open-source flight control firmware, actively maintained by the Dronecode Foundation. Like ArduPilot, it runs on Pixhawk-compatible hardware. PX4 records flight logs in a binary format called ULog (.ulg, .ulog), which is structurally different from ArduPilot's DataFlash .bin format.
LogHat natively supports PX4 ULog files. Upload a .ulg file directly — no conversion required.
.bin Log File
A .bin file is a binary-encoded flight log produced by ArduPilot's DataFlash logging system. It is written to the SD card in the flight controller during every powered session. The file contains dense, structured telemetry data including attitude (roll, pitch, yaw), GPS position, barometric altitude, battery state, motor outputs, navigation state, and system messages — all logged at configurable rates (often 25–400Hz).
The .bin format is not human-readable directly. Software tools such as Mission Planner, QGroundControl, MAVExplorer, or LogHat are required to parse and interpret the data.
EKF (Extended Kalman Filter)
The Extended Kalman Filter (EKF) is the sensor fusion algorithm at the heart of ArduPilot's navigation system. It continuously combines data from multiple onboard sensors — gyroscopes, accelerometers, barometer, GPS receiver, and compass — to compute a reliable estimate of the drone's position, velocity, and orientation in three-dimensional space.
The EKF operates by maintaining an internal model of the aircraft's physics and using incoming sensor measurements to update and correct that model. When sensors disagree (for example, when GPS position conflicts with accelerometer-derived position estimates), the EKF uses statistical weighting to determine which source to trust more.
An EKF failure or EKF health warning in a log indicates that the filter lost confidence in its state estimate — which can cause unexpected flight behavior including position drift, altitude errors, and in severe cases, loss of control.
Common causes of EKF failures: compass interference, GPS multipath, excessive vibration corrupting accelerometer data, sudden GPS signal loss.
Vibration FFT
FFT stands for Fast Fourier Transform, a mathematical algorithm that converts a time-domain signal (vibration amplitude over time) into the frequency domain, revealing which frequencies contain the most vibration energy.
In the context of drone flight logs, vibration FFT analysis is used to identify the dominant vibration frequencies in the airframe. This is useful for diagnosing the source of vibration:
- Motor/propeller fundamental frequencies — Vibration at the motor rotation frequency indicates propeller imbalance or motor bearing issues.
- Frame resonance — Vibration that matches natural structural frequencies of the frame, which can be amplified under certain throttle settings.
- Harmonic frequencies — Multiples of the fundamental frequency, which may indicate nonlinear vibration sources.
While LogHat reports vibration amplitude metrics (VibeX, VibeY, VibeZ), full FFT analysis is performed by the deeper engineering analysis engine and surfaced in the forensic PDF report.
GPS HDOP / VDOP
HDOP (Horizontal Dilution of Precision) and VDOP (Vertical Dilution of Precision) are dimensionless numbers that quantify the accuracy of GPS position estimates based on the geometric arrangement of visible satellites.
DOP values are multipliers on GPS measurement error. A HDOP of 1.0 means the horizontal GPS error equals the base measurement uncertainty. A HDOP of 3.0 means horizontal GPS error is three times the base uncertainty.
| DOP Value | Quality | |---|---| | 1.0 | Ideal | | 1.0–2.0 | Excellent to acceptable | | 2.0–5.0 | Moderate — position accuracy degraded | | Above 5.0 | Poor — should not fly in autonomous modes |
HDOP is primarily affected by the number of visible satellites and their spread across the sky. Satellites bunched together in one part of the sky produce higher DOP than satellites spread evenly. Flying in urban canyons or under heavy tree cover reduces satellite count and worsens DOP.
LogHat applies a health score deduction when HDOP exceeds 2.0 during any sustained portion of the flight.
Battery Sag / Voltage Drop
Battery sag (also called voltage drop under load) is the reduction in a battery's terminal voltage that occurs when current is drawn from it during high-throttle maneuvers or hover under heavy payload.
All lithium-based batteries have internal resistance. When current flows, this internal resistance causes a voltage drop proportional to the current drawn. As a battery ages or if cells are damaged, internal resistance increases and voltage sag worsens.
High battery sag is significant for two reasons:
- Performance: A sagging battery delivers less power to motors, limiting the drone's ability to respond to rapid attitude corrections.
- Low-voltage failsafe: If the sag is severe enough, the battery voltage may momentarily drop below the failsafe threshold during a high-throttle maneuver, triggering an emergency landing even if the battery's average charge level is adequate.
In LogHat's analysis, battery sag is calculated as the difference between the mean resting voltage (at low current) and the minimum voltage recorded during peak current events.
Health Score
The health score is LogHat's unified assessment of a flight's safety and airworthiness, expressed as a number from 0 to 100. It is calculated by the analysis engine after your log is processed and represents a weighted combination of EKF health, vibration levels, GPS signal quality, and battery behavior.
A score of 100 indicates a flight where all analyzed parameters were within normal operating ranges. Deductions are applied for each detected anomaly, with more severe issues resulting in larger deductions. The score is accompanied by a status badge:
- Passed (80–100): No significant issues.
- Warning (50–79): One or more parameters require attention.
- Grounded (0–49): Critical issues; review before next flight.
See AI Flight Health Score for the full scoring criteria.
DataFlash
DataFlash is the name of ArduPilot's onboard logging subsystem. It records binary log files to an SD card at high frequency during flight. DataFlash logs are organized as a sequence of typed messages (ATT for attitude, GPS for position, BAT for battery, BARO for barometric altitude, and many more), each with a schema defined by FMT (format) messages at the start of the log file.
The .bin file extension is the standard output of the DataFlash logging subsystem and the primary file format supported by LogHat.
Failsafe
A failsafe is an automated safety action taken by the flight controller when a critical condition is detected and the pilot is unable to safely override. Common failsafe triggers include:
- Radio/RC signal loss — The controller cannot receive input from the pilot
- Battery low-voltage — The battery voltage drops below a configured threshold
- GPS glitch — GPS position estimate becomes unreliable
- GeoFence breach — The aircraft exits a configured safe operating area
- EKF failure — The navigation state estimator detects a critical inconsistency
Typical failsafe actions include Return to Home (RTH) or immediate landing. Failsafe events are recorded as ERR messages in the flight log and are identified and listed in the critical alerts section of the LogHat PDF report.
IMU Clip
An IMU clip (Inertial Measurement Unit clip) occurs when the vibration measured by the accelerometer exceeds the maximum measurable range of the sensor hardware. When this happens, the sensor output is "clipped" at the hardware maximum value, meaning the actual vibration level was even higher than what was recorded.
IMU clips are a serious indicator of extreme vibration. Even a small number of clip events in a flight log indicates that the vibration environment at the flight controller's mounting point exceeds the sensor's measurement range, which can corrupt navigation estimates significantly. LogHat counts IMU clip events and flags any clips as a Warning or Grounded condition.
Return to Documentation Home