PPK vs RTK for drone survey: which should you use and when?

1. Why this question actually matters

Every professional drone survey operator eventually faces the RTK vs PPK decision — often under pressure, on-site, with a client waiting for results. The answer they find in most online content is a polished non-answer: both methods deliver centimetre-level accuracy, it depends on your project, here is a comparison table. That is technically correct and operationally useless.

The real question is not which technology is better in a vacuum. It is: given your specific environment, equipment, infrastructure, and deliverable requirement — which method is actually going to give you reliable, defensible survey data on this mission? The answer changes depending on whether you are mapping a construction site in Georgetown with 4G coverage and a clear sky, or flying a stockpile survey 200 km into Guyana’s interior with no CORS network, intermittent radio contact, and a charter flight window that does not accommodate re-flights.

2. What RTK and PPK actually are — the technical foundation

Both RTK and PPK are methods for achieving centimetre-level GNSS positioning accuracy. Standard GPS — the kind built into consumer electronics — delivers accuracy in the range of 1–5 metres outdoors. Professional survey work requires accuracy of 1–3 cm horizontal and 2–4 cm vertical. The gap between those two figures is bridged by carrier-phase correction techniques, of which RTK and PPK are the two dominant variants in UAV operations.

Both methods work on the same underlying principle: a base station at a known, fixed location on the ground records GNSS satellite observations simultaneously with the drone in flight. The base station’s known position allows calculation of the exact error affecting the GNSS signal at that moment — atmospheric delays, satellite clock errors, orbital errors — and those corrections are applied to the drone’s raw position data to produce centimetre-accurate geotagged image coordinates. The difference between RTK and PPK is entirely about when those corrections are applied.

2.1 RTK — Real-Time Kinematic

In RTK, the base station transmits its correction data to the drone in real time during the flight, via radio link from the base station to the remote controller, and from the controller to the aircraft. The drone’s GNSS receiver fuses these corrections on the fly, producing corrected image geotags as each photo is taken.

This requires four simultaneous communication links to be maintained at all times:

• Satellite signals to the drone’s GNSS receiver
• Satellite signals to the base station
• Base station to remote controller (radio/internet correction stream)
• Remote controller to drone (link between correction data and onboard receiver)

If any of these links degrades or breaks, the RTK solution downgrades from ‘fixed’ to ‘float’ or drops to standalone GPS entirely. This is the critical operational vulnerability of RTK, and it is far more common in field conditions than most pre-sale demonstrations suggest.

2.2 PPK — Post-Processed Kinematic

In PPK, the base station and the drone each independently record their raw GNSS observation data throughout the flight. No real-time communication link between them is required during the mission. After the flight, the two raw data files are combined in post-processing software, which resolves the carrier-phase ambiguities and applies corrections retroactively.

The critical structural advantage of PPK is that it decouples data collection from correction. If the drone’s radio link to the base station is interrupted mid-flight, the raw GNSS logs from both receivers continue uninterrupted. Nothing is lost. The corrections are applied in the office, to complete data, not to a fragmented real-time stream.

2.3 RTK fix vs float — the number most reviews skip

The single most important technical distinction that operators need to understand is the difference between an RTK fixed solution and an RTK float solution. It determines whether your RTK data is survey-grade or nearly useless, and the distinction is invisible in the imagery itself — it only appears in the flight log.

A float solution can have errors of decimetres to over a metre. It is not centimetre-accurate data, regardless of what the drone’s display shows. When the solution transitions from fixed to floating conditions, errors can reach decimetre or even metre level. There is additionally a documented phenomenon called a ‘false fix’, where the receiver reports a fixed solution but the underlying ambiguity resolution has failed — producing errors at metre scale while displaying centimetre-level confidence.

3. Accuracy numbers — what the research actually shows

Both RTK and PPK, when properly executed, deliver comparable final accuracy: 1–2 cm horizontal and 2–4 cm vertical under good conditions with a fixed solution. Post-processed volumetric accuracy for stockpile surveys is typically within 1–3% of ground-truth methods. These numbers assume correct execution.

3.1 Baseline length — the distance factor

RTK accuracy degrades with increasing distance between the drone and the base station, following a predictable formula: approximately 0.5 mm of additional horizontal error per kilometre of baseline (0.5 ppm), and 1.0 mm of additional vertical error per kilometre (1.0 ppm).

PPK is substantially more tolerant of baseline length because the corrections are resolved post-flight using full observation data and more sophisticated atmospheric modelling. PPK maintains survey-grade accuracy at baselines up to 100 km under good satellite geometry — a practical distinction that matters enormously for Guyana interior operations.

3.2 What happens when RTK signal is interrupted

A one-second interruption does not just degrade that one second of data. RTK requires re-initialisation to recover a fixed solution after a dropout — typically 30–90 seconds in good open-sky conditions. During re-initialisation the system is in float mode. In obstructed environments, re-initialisation may not complete before the next interruption.

4. When to use RTK — the conditions that actually support it

RTK is the right primary method when the following conditions are genuinely met, not just assumed:

• Clear sky above the survey area — less than 20% obstruction above 15° elevation angle throughout the entire mission, not just at the launch point
• Stable communication link — radio link between base and controller confirmed at all points in the flight plan before takeoff
• Short baseline — base station within 10 km of the survey area for reliable fixed solution in tropical atmospheric conditions
• Confirmed RTK fix before takeoff — do not begin the mapping mission until the aircraft has reported fixed solution status for at least 60 continuous seconds
• Real-time results required — the client or workflow needs corrected data immediately on landing
• CORS network available — or own base station on a known point within baseline limits

4.1 RTK on active construction sites

For construction earthworks progress monitoring, RTK offers a genuine workflow advantage: the site manager can review preliminary data on the remote controller immediately after landing. For clients who need to make same-day decisions — continue grading, pause for survey, re-sequence earthworks — the immediacy of RTK data has real operational value.

4.2 RTK for smaller, well-defined sites

Sites under approximately 50 hectares, well within radio range of the base station and with good sky visibility, are the natural RTK use case. The entire site is covered before atmospheric conditions or link stability have opportunity to degrade.

5. When to use PPK — the cases where it is the superior method

PPK is not a fallback for when RTK fails. It is the technically superior method in a significant range of professional survey scenarios.

5.1 Remote operations without CORS infrastructure

Guyana has no national CORS network. This is not unusual — the majority of developing-world mining, O&G, and construction markets operate without continuous GNSS reference infrastructure. In these environments, PPK removes the requirement for a stable real-time link. The base logs continuously and independently. If the radio link drops, the logging continues. In Guyana interior operations, where radio link reliability over forest clearings is unpredictable and re-flight is not an option, PPK is the only workflow that provides reliable survey-grade results.

5.2 Large-area surveys and extended baselines

When a survey area is large enough that the drone is consistently flying at the edge of reliable radio range, PPK is structurally more reliable. The drone and base station operate independently; the only requirement is that both are logging continuously.

5.3 High-value missions where re-flight is not viable

For mining clients paying for a helicopter mobilisation to an interior site, or for O&G clients with a narrow access window, a failed RTK dataset is not just inconvenient — it is a contractual and commercial problem. PPK eliminates the single point of failure that causes RTK datasets to fail: the real-time communication link.

5.4 PPK as the QA backstop for RTK data

On the DJI Matrice 4E, the aircraft logs raw GNSS observation data simultaneously with RTK correction reception. The DJI RTK 3 base station logs raw observation data continuously. This means that every RTK mission is also, automatically, a PPK mission in waiting — as long as the operator collects the base station log file before departing the site.

6. The workflows — step by step

6.1 RTK workflow — DJI M4E + DJI RTK 3

6.2 PPK workflow — DJI M4E + DJI Terra

7. Full RTK vs PPK comparison

8. The Guyana context — operating without CORS infrastructure

Guyana has no national continuously operating reference station (CORS) network. This is relevant not just as a local detail but as a proxy for a large class of operating environments: resource-sector markets in the developing world where the survey infrastructure that operators in North America or Europe take for granted does not exist.

8.1 Coastal belt operations

In accessible coastal belt operations — construction sites on the East Bank Demerara, infrastructure surveys around Georgetown — RTK with the DJI RTK 3 on a known point is viable as a primary method. Sky view is generally good on flat coastal terrain. Baselines are typically under 5 km. RTK fix acquisition is fast and maintenance is reliable.

8.2 Interior operations — mining and O&G sector

Interior operations present the full range of conditions that make RTK unreliable as a primary method:

• No CORS network means the base station position is established by averaging or by occupation of a previously surveyed benchmark
• Forest clearings and riverside airstrips mean partial sky obstruction — RTK fix maintenance over sites adjacent to tall secondary growth is unpredictable
• Charter flights mean mission windows are fixed — a failed RTK dataset cannot be recovered if raw GNSS logs were not collected
• Radio link over forest terrain is less reliable than over flat coastal land

9. Where GCPs fit into this

9.1 GCPs alongside RTK/PPK

RTK and PPK both reduce the number of GCPs required — from the 8–12 typical for standard photogrammetry to 0–2 for validation checkpoints. The distinction matters:

• GCPs for accuracy correction: Placed before the flight, their coordinates correct the photogrammetric model during processing. Required when base station position accuracy is uncertain.
• Checkpoints for accuracy validation: Placed before the flight but not used in processing. Their known coordinates are compared against the processed model to generate an independent accuracy statement.

For mining stockpile volumetrics and construction earthworks, the minimum professional standard is 1–2 checkpoints per survey area, measured independently with GNSS rover or total station.

9.2 GCPs without RTK/PPK

Standard photogrammetry without RTK or PPK relies entirely on GCPs — typically 6–8 well-distributed points. The field time to place, measure, and recover them on every mission is significant. RTK and PPK exist precisely to eliminate this operational burden.

10. Decision guide — which method for your mission

11. Common mistakes that compromise survey accuracy

1. Declaring survey-grade results from RTK data without checking the flight log for float periods. The RTK status display during flight is not a sufficient record — download and review the full flight log.
2. Forgetting to collect the base station log file before leaving the site. Once overwritten on the next mission, the PPK backstop is gone permanently.
3. Setting the base at an unknown point and assuming averaged position is good enough. Averaging improves relative accuracy but absolute position may be off by metres. For audit-quality mining inventory, this matters.
4. Using an incorrect coordinate reference system. Guyana operations should use WGS84 / UTM Zone 21N. Mixing CRS silently shifts the entire dataset.
5. Not verifying PPKRAW.bin exists before departing the site. Check on the drone’s storage via the RC Plus 2 before powering down.
6. Assuming a second flight corrects a bad first flight. If the base position was wrong, the error propagates to both datasets equally.

12. Equipment for this workflow

13. Frequently asked questions

14. Summary

RTK and PPK achieve the same final accuracy when executed correctly. The difference is in reliability and execution requirements, not in the physics of the correction.

RTK is faster to result, better suited to accessible sites with good sky view and stable communications, and delivers immediate on-site feedback. PPK is more robust, tolerates communication failures and longer baselines, and is the standard production method for any operation where re-flight is expensive or impossible.