GCP vs RTK vs PPK Aerial Mapping: Which is Better?

Posted on

Aerial mapping is one of the fields that has the potential of being completely revolutionized by drone technology. Drones allow surveyors to map huge areas quickly and accurately without exposing themselves to the hazards of traffic, rough terrain, or wildlife. The field of drone mapping has evolved so much in the last few years that different methods and technologies for carrying out a mapping survey have already been developed.

From the traditional use of Ground Control Points (GCP) to drones capable of Real-Time Kinematic (RTK) and Post-Processed Kinematic (PPK) technology, the choice of which method to use may be overwhelming for a beginner drone surveyor. In this article, we discuss how GCPs, RTK, and PPK differ from each other, and compare their accuracies when used in conducting drone surveys.

What is GPS anyway?

At the very core of discussing the different methods for doing drone surveys is GPS technology. The Global Positioning System (GPS) is a technology for determining the exact location of objects using a satellite-based navigation system. At any given point, there are at least 24 different GPS satellites positioned in six orbital planes. Each satellite transmits a unique signal which a GPS receiver decodes to determine the receiver’s distance from the satellite. The exact location of a GPS receiver is determined by receiving information from several satellites and applying a process of trilateration – basically, a more sophisticated version of triangulation.

Although a GPS receiver only needs to receive a signal from 3 satellites to estimate its exact location, most receivers will track 8 or more satellites to improve locational accuracy. In addition to tracking the location of a GPS receiver in 2D space, the use of more GPS satellites allows for tracking in the 3D space – that is altitude in addition to longitude and latitude. By successive measurements of a receiver’s location, a GPS receiver can calculate for an object’s speed, direction, and distance travelled.

Nowadays, GPS technology is no longer limited to specialized handheld receivers. Practically all smartphones have GPS capability, and the technology has also been integrated into smartwatches, cars, bikes, and scooters. GPS tracking can be used to measure a running route, provide road navigation, or map a golf course.

Since a GPS receiver’s accuracy is dependent on the quality of signals it receives from the satellites, its calculated location will have an inherent degree of error. Satellite signals can be degraded as they pass through the atmosphere and will be bounced around large structures such as buildings and tress before they reach the receiver. These factors, among several others, affect the locational accuracy of GPS receivers.

Most consumer-level GPS receivers are accurate between 10 to 20 meters, which is usually enough for general navigation. Without the aid of ground-level correction, even professional-grade GPS receivers can achieve an accuracy of only 2 to 5 meters. A positioning technique that uses Real-Time Kinematic (RTK) technology can be done to achieve centimeter-level accuracy via “ambiguity resolution” – a process that involves reduction and removal of common errors. This development has greatly enhanced the output of mapping surveys that rely on GPS technology, as we shall discuss in greater detail below.

What are GCPs?

Ground Control Point (GCPs) are marked points on the ground that are strategically placed around the survey area. These points are identified with large and highly visible markers, and their locations are accurately determined using RTK or PPK-enabled GPS receivers. By using the accurate locations of the GCPs, reproduced maps can be “calibrated” for better absolute accuracy. The use of GCPs has been an established practice for mapping applications that rely on a high level of global accuracy, such as in land title surveys, as-built surveys, and maps that need to be integrated with other geospatial data.

For best results, GCPs are often distributed along the edge or corners of the survey area. Whatever geometry a survey area may take, it is often recommended to have a GCP near the middle of the area of interest. A minimum of 5 GCPs is generally recommended for a survey of any size or objective. Larger areas may benefit from having more GCPs, but there is little advantage in having more than 10 GCPs in terms of accuracy.

The use of GCPs represent the more traditional practices of the field of mapping surveys. Taking ground-level measurements is one of the most laborious and time-consuming steps of a survey, but it is recognized as a necessary sacrifice to ensure a map’s consistency with a global geodetic coordinate system. By anchoring the results of an aerial survey to a “ground truth”, datasets and the resulting maps and models gain an unequaled level of robustness and reliability.

Establishing GCPs is a crucial step when conducting a drone-survey using a non-RTK enabled drone.  As with standard GPS receivers, the GPS position of a drone can only be determined to an accuracy of around 10 meters. Compounding the inherent error of the GPS receiver, geotagging of the images capture by a drone’s onboard camera also must consider the camera angle, the timing of the camera’s trigger, and the rotational position of the drone. The use of GCPs can reduce the error margin of survey data to 3 meters or less.

What is RTK and PPK and how are they different?

More recent drones designed for mapping surveys boast of their integration with Real-Time Kinematic (RTK) modules. Manufacturers claim that these drones can achieve centimeter-level accuracy for maps and models and reduce the need for GCPs. What is RTK and how does it work?

RTK-enabled drones use differential GPS measurements to improve accuracy. In a nut shell, this means that an RTK workflow involves measuring the location of a base station. You can think of this base station as a single GCP that constantly provides correction and calibration of the drone’s locational data. Successive GPS measurements at the base stations are paired with the GPS measurements made by the drone. This provides a mechanism for the reduction and elimination of the errors common between the two measurements. These corrections are applied real-time, requiring the drone and the base stations to communicate with each other constantly throughout the survey.

However, the limitation of using a purely RTK-enabled drone is its need to maintain a constant and uninterrupted connection with the base station. This is not possible in all survey areas, as large structures such as buildings, trees, and mountains can easily interfere with the signal exchange between the two components. Should they get disconnected, an RTK-enabled will tend to revert to standalone mode or produce RTK-float data. This greatly reduces location accuracy, defeating the purpose of using an RTK-enabled drone in the first place.

By using a Post-Processed Kinematic (PPK) workflow, this limitation can be bypassed. Instead of relying on real-time data correction, a PPK drone collects location data independently. At the same time, the base station collects its own set of locational data, albeit at a much higher level of accuracy. After the survey, the timestamps of the two datasets are used to reconcile the data and make post-process adjustments. The less accurate data of the drone’s onboard GPS is corrected using the more accurate base station data, resulting in more precise geotags of aerial imagery or other survey data.

Drones that process data using RTK or PPK do not have different sets of hardware. Instead, a PPK drone merely uses a different processing workflow. Data collected by an RTK drone can be corrected using a PPK process, but it will involve a lot of customization. For best results, a PPK-enabled drone is still recommended to collect PPK-grade data.

PPK vs RTK?

Given the two workflows for differential positioning technology, is there a benefit to using one over the other?

PPK has the advantage of allowing for more flexibility in terms of the actual survey execution. Not having to rely on maintaining a constant communications link means that the survey can be done even in areas where there are a lot of potential obstacles. Moreover, data collected using a PPK drone can be referred to previous or future data for comparison or contrast. This provides a much higher level of reliability and robustness of data.

PPK surveys are especially appropriate for urban areas where buildings and houses can render RTK surveys severely ineffective. In a study conducted by Pix4D, an RTK survey conducted in an urban area achieved only a 71% rate of well-corrected camera positions. The remaining 29% were not corrected real-time, and were classified as “RTK-float” Although the RTK survey achieved an impressive Z-dimension accuracy with an error of only 8.1 cm, PPK processing of RTK-float data resulted in this error being reduced to 6.7 cm.

In conclusion, a drone with PPK workflow capabilities is vastly superior to an exclusively RTK-enabled drone. It is more reliable and robust in terms of data handling and provides more flexibility in where a mapping survey can be conducted.

GCPs vs. RTK/PPK?

Having established the difference between RTK and PPK, we will now assess how this new mapping technology compares to the traditional method of using GCPs.

In terms of survey execution, the use of an RTK or PPK-enabled drone has the potential of reducing the number of GCPs needed to conduct an accurate mapping survey. In most cases, a survey using an RTK-enabled drone only needs 1 or 2 GCPs. This is a huge improvement to the minimum of 5 GCPs recommended for drones with standard GPS receivers. By eliminating the legwork needed to take ground-level measurements at GCPs, the physical effort and time needed to conduct a mapping survey can be greatly reduced by using an RTK-enabled drone.

The upgrade in accuracy afforded by using an RTK-enabled drone is exhibited best by conducting a survey in an open field, as conducted by Pix4D in their study. By flying in an area that is free from any obstacles, the RTK drone was able to receive signal from the base station for 99% of the data points. This resulted in an error of only 2.7 cm. A survey conducted in the same area using a standalone drone and well-distributed GCPs only succeeded in achieving errors of 36 cm in the X-axis and up to 3.5 meters in the Z-axis.

However, GCPs can still be every useful in survey areas that are full of obstacles. In the same urban area mentioned in the previous section, a survey using an RTK-enabled drone and well-distributed GCPs achieved error values of only 4.8 cm. In fact, correction using GCPs resulted in slightly better accuracy than correction using PPK. Considering the 6.7-cm error of the PPK-corrected data, the results were very close.

Which technology to use?

When comparing RTK drones and PPK drones, the PPK drones easily come out as the better choice. Although real-time data correction sounds appealing, it is not a process that can be reliably done in real-world conditions. Urban or heavily vegetated areas that are dense with potential obstacles will prevent the drone from getting a fixed signal from the base station, neutralizing the accuracy-enhancing benefits of RTK drones.

The post-flight data processing of PPK drones allows them to collect data reliably in more diverse environments. Moreover, data handing using a PPK workflow gives the surveyor the opportunity to compare present survey data from past and future datasets, increasing its reliability and robustness.

However, this does not mean that GCPs no longer have a place in the modern landscape of mapping surveys. As many surveyors will say, there is no replacement for ground truth data. Even the most sophisticated processing algorithms cannot equal the level of reliability of data collected at the ground level. For most mapping applications, GCPs are still recommended to produce a very defensible and undeniably valid output.

Exceptionally large survey areas can still benefit from using GCPs to maintain a high level of accuracy even in sections of the survey area that are far from the base station. In the absence of PPK capabilities, data collected using an RTK drone can be corrected using GCP measurements to greatly enhance its accuracy.

The best solution that surveyors have come up with is to combine established GCP practices with the technological advantage of RTK. Thus, a survey conducted using an RTK-enabled drone can be carried out with just 1 or 2 GCPs.

Is RTK/PPK worth it?

One major hurdle that is limiting the wide adoption or RTK or PPK drones is the high cost of initial investment associated with the technology. Aside from RTK-equipped drones being more expensive than those with standard GPS receivers, surveyors also need to purchase an RTK-compatible base station. This whole bundle can be more than 5 times more expensive than a typical camera drone with GPS capabilities.

However, conducting mapping surveys using a full complement of GCPs involves spending hours or days of taking ground-level measurements. This corresponds to additional manhours and logistical expenses. There is also considerable time spent on planning where to place GCPs, a task that may prove unproductive should a GCP turn out to be inaccessible do to terrain limitations.

All things considered, purchasing RTK technology is a matter of spending on a one-time investment versus the recurring costs associated with maintaining the traditional practices of establishing multiple GCPs. For businesses that will operate for the long term, the one-time investment for an RTK drone is the more financially sound decision.

RTK capabilities also allow surveyors to upgrade their capabilities with enhanced accuracy and shorter turnaround times. In a field that is becoming more and more competitive, this is a business advantage that may prove to be invaluable in the long run.

Final thoughts

The development of RTK and PPK technology has been a largely welcomed upgrade to survey drones. It represents a large step forward in terms of mapping survey practices, a field that already been significantly revolutionized by drone technology. With RTK technology, more accurate maps and models can be produced at a fraction of the effort traditional methods required.

Currently, the high costs associated with upgrading to RTK technology has proven to be a limitation that the field of drone mapping is yet to overcome. As the technology continues to mature, it has the opportunity of becoming more affordable. In the near future, we expect more and more surveyors and mapping firms to recognize upgrading to RTK drones as a good business decision.