Satellite navigation is useful, but a UAV should not treat GNSS as its only source of position. Signals can be obstructed, reflected, degraded, or unavailable. In a GPS-denied or GPS-degraded environment, safe navigation depends on estimating what the vehicle can still know, how uncertainty is growing, and whether the remaining sensors support the intended operation.
What an inertial navigation system contributes
An INS uses inertial measurement unit (IMU) data—typically accelerometers and gyroscopes—to estimate changes in attitude, velocity, and position. Because small sensor biases are integrated over time, inertial position uncertainty grows during extended operation without external corrections. A high-grade IMU can reduce some errors, but it cannot eliminate drift or substitute for independent position observations indefinitely.
Sensor fusion and estimator health
An estimator such as an extended Kalman filter can combine measurements with different error patterns and update rates. Inertial data provides continuity; GNSS can correct long-term position drift when healthy; barometers, range sensors, cameras, and other references may contribute where suitable. The estimator must account for measurement timing, calibration, vibration, sensor alignment, and uncertainty.
More sensors do not automatically mean a more reliable estimate. A faulty, delayed, or misaligned measurement can make navigation worse if accepted without validation. Monitor estimator innovations, sensor health, quality flags, and failsafe state. Flight behavior under degraded navigation should be defined and tested for the specific aircraft and operating approval.
Optical flow and local positioning
Optical-flow sensors estimate apparent movement of image features across the camera field of view. With suitable downward-looking imagery and a valid distance measurement—often from a rangefinder—the system can estimate horizontal motion relative to the surface. It is useful for local position hold in conditions with adequate texture, lighting, range, and sensor alignment.
Optical flow is not a universal GPS replacement. Featureless surfaces, low light, excessive altitude, rapid motion, dust, water reflections, or an unreliable distance reading can reduce its usefulness. It generally provides local relative motion, not a globally referenced geographic position. The estimator and flight mode must be configured for the sensor combination actually installed.
High-precision and resilient GNSS
Multi-constellation receivers, dual antennas, and correction services such as RTK can improve availability, heading, or position accuracy in supported installations. They still depend on usable satellite signals and correct antenna placement. Redundant receivers can help detect disagreement or provide a fallback, but they do not make a system immune to common interference or shared installation faults.
For interference resilience, select equipment with documented interference monitoring and receiver integrity features, follow the receiver and autopilot manufacturer’s integration guidance, and assess the antenna, cabling, filtering, and electromagnetic environment. Anti-jamming claims should be evaluated against independent specifications and the applicable regulatory and spectrum rules. Do not transmit interference; GNSS jamming is unlawful in many jurisdictions and can endanger other users.
Design for graceful degradation
- Identify which navigation sources are available for each route, altitude, and environment.
- Calibrate and mechanically isolate sensors according to manufacturer instructions.
- Set health checks and failsafes based on estimator confidence, not simply a GPS fix indicator.
- Test loss and recovery of navigation inputs in simulation and controlled conditions before operational use.
- Define conservative actions—such as hold, return, or land—only where the aircraft can safely perform them with the remaining references.
Robust GPS-denied navigation is a system property: sensors, estimator, airframe, software configuration, and operating procedure must work together. High-precision GNSS can improve the solution when signals are trustworthy; inertial and local sensing provide continuity when they are not, with uncertainty managed throughout.
Further reading
- ArduPilot: GPS and compass positioning, blending, and interference handling
- ArduPilot: Optical-flow sensor setup and testing
- ArduPilot: EKF state estimation overview