RTK

A navigation device in a car determines its location using signals from satellites. This location reading is normally accurate to within about three to ten meters. That’s sufficient for road navigation, but not for a tractor that needs to hit a seed row precisely. RTK is an additional technique that increases accuracy to one to two centimeters. The abbreviation stands for Real Time Kinematic: the correction happens during ongoing operation, while the device is moving. To achieve this, in addition to its own receiver, a second station is needed whose exact position is already known.

Why centimeters instead of meters make the difference

Many machines simply cannot do anything useful with an error of five meters. A self-steering tractor with this level of inaccuracy would till entire lanes twice or skip strips altogether. With RTK, it drives rows that lie side by side accurate to the centimeter. This saves seed, fertilizer, and diesel, and the driver doesn’t have to constantly make corrections.

The situation is similar on construction sites. An excavator digging a pit to an exact depth today often works with RTK-assisted control. Surveyors use it to replace work steps that used to require a tripod, a measuring rod, and a second person. Drones mapping fields or buildings also need centimeter-accurate positions, otherwise the individual aerial images won’t line up properly afterward.

There’s a common misunderstanding: RTK doesn’t make the map more accurate, it makes the device’s position on the map more accurate. If the underlying map itself is inaccurate, RTK won’t help with that. The two sources of error are independent of one another.

Base station, rover, and the trick with the carrier wave

The inaccuracy of ordinary satellite positioning arises mainly in the atmosphere. The signals are slightly delayed on their way to Earth, and this delay fluctuates. The crucial point: two receivers standing only a few kilometers apart experience almost the same error. This is exactly what RTK exploits.

To do this, a base station is set up at a precisely surveyed point. It receives the same satellite signals as the mobile device, known as the rover. Because the base station knows its true position, it can calculate the current error. It transmits these correction values to the rover via mobile network or radio modem, and the rover immediately incorporates them into its own calculation.

There’s a second trick involved. Normal receivers evaluate the modulated code of the satellite signal, whereas RTK additionally measures the carrier wave itself. Its wavelength is only about 19 centimeters, which allows for much finer measurement. However, the receiver must first work out how many whole waves lie between it and the satellite. This lock-in process takes seconds to minutes and is lost if a bridge or a tree blocks the line of sight.

From the surveying crew to the smartphone

In practice, hardly anyone sets up their own base station anymore. Instead, there are nationwide correction services with permanently installed reference stations. In Germany, the national survey authority operates the SAPOS network, and commercial providers exist alongside it. The user receives the correction data via a mobile connection and pays a fee for it.

In product descriptions, RTK mainly appears in connection with surveying equipment, agricultural machinery, and higher-end drones. Newer-generation robotic lawn mowers also advertise it: they no longer need a boundary wire in the lawn because they know their own position. In autonomous vehicles, RTK is one component among several, complemented by cameras and laser scanners.

RTK should be distinguished from PPP and from SBAS. PPP works without a nearby base station, but takes considerably longer to reach full accuracy. SBAS, known in Europe as EGNOS, improves position only to about one meter, but is free of charge. Which method makes sense depends on how accurate, how fast, and how expensive it’s allowed to be.