During a recent flight test, the drone's camera was physically obstructed for the whole of the flight yet it still managed to land at the correct target coordinates. The positioning was obtained entirely from the flight telemetry and processed via a cloud-based service, without any need for a GPS signal or visual input at all. This is no mere novelty — it is a direct solution to a problem that operators are becoming increasingly aware of, namely, what happens to your data when the GNSS signal is simply not reliable during the flight that you actually need to make.
Jamming vs. Spoofing, Two Different Failure Modes
They are grouped together but constitute different types of failure. Jamming is direct in that it overloads the GNSS frequencies with radio-frequency noise, thereby preventing the receiver from being able to acquire or track any signal. Spoofing is more subtle since it sends out counterfeit signals which have been designed to appear genuine, causing the receiver to calculate a false position instead of no position at all. GPS signals have to travel thousands of kilometres from orbit and reach the Earth's surface at extremely low power, so that even a relatively small ground-based jammer can provide service denial over a surprisingly large area.
Why RTK's Biggest Weak Point Isn't the Drone
Most high-precision drone operation procedures use RTK corrections to achieve an accuracy of a few centimetres, and there is one particular weakness that is important to be aware of: the fixed reference station can itself be jammed or spoofed. Research into this exact method of attack has demonstrated that by altering the position-navigation-time solution of the reference station, degraded accuracy or completely false corrections are transmitted to all rovers connected to it — which means that the weak link in your accuracy chain is usually not the drone's receiver at all, but rather the ground-based infrastructure providing the corrections.
What Actually Breaks When This Happens
The main worry is position drift, but this isn't the entire issue. Jamming which interferes with carrier-phase continuity also impairs attitude determination—that is, the drone's knowledge of its own heading and orientation—and this relates to a flight-control problem, not merely a data-quality one. Spoofing can go even further: studies have recorded spoofing attacks that were specifically designed to cause false heading deviations so severe as to affect the drone's flight behaviour rather than just corrupting the coordinates afterwards.
What Backup Approaches Exist Today
The conventional method is dead reckoning using an inertial measurement unit (IMU), whereby the vehicle's position is estimated on the basis of onboard motion sensors when the GNSS signal fails, although the accuracy of this method decreases the longer the aircraft continues to fly without a GNSS correction. The more recent approach currently attracting attention, on the other hand, follows a completely different path: it involves cloud-based positioning-as-a-service, which determines the location from flight telemetry alone instead of requiring additional onboard hardware. In the flight trial referred to earlier, this method was able to calculate the correct target coordinates even when the aircraft's camera was covered for the entire flight, proving that the positioning calculation did not rely on any single onboard sensor functioning properly — thus offering a considerably different kind of resilience compared with merely adding a backup GPS chip.
What This Means for Survey Operators
If you are carrying out GIS and drone survey work in regions where GNSS interference is even possible—such as near airports, at industrial sites, or any place where spoofing or jamming has been reported—it is worthwhile getting into the habit of checking the fix quality metadata in your drone imaging and processing output before accepting it at face value, rather than simply assuming that an orthomosaic has been accurately produced because it has been rendered without errors. A flight which has had a short RTK dropout can still result in a visually clean output even if it has substantially degraded positional accuracy underneath. Understanding the various types of failure—such as jamming and spoofing, vulnerability of the reference station, and attitude degradation—is what makes the difference between spotting the problem before it gets to the client and discovering it afterwards.
