Basics and Limitations of RTK (Real-Time Kinematic) Positioning

RTK positioning is a technique that delivers centimetre-level accuracy by correcting GNSS errors in real time. It uses two GNSS receivers simultaneously: a base station installed at a point of known coordinates and a rover that moves while measuring its position. By analysing the phase differences of the satellite signals received by both units, RTK cancels out common error sources—such as satellite-clock offsets and ionospheric or tropospheric delays—keeping positioning errors within just a few centimetres.

Put simply, the base station calculates the error in the satellite data and transmits that information to the rover, which applies the corrections to achieve high accuracy.

However, RTK still depends on GNSS signals, just like standard GPS. In environments where radio signals are weak or obstructed, RTK accuracy cannot be maintained. If satellite signals are lost, the base station cannot send correction data and positioning becomes unreliable. Urban canyons, for instance, often suffer from multipath reflections and signal blockages caused by tall buildings, while tunnels completely block satellite signals, forcing RTK to halt. In such situations, relying solely on RTK leads to “no-fix” conditions and interrupted positioning.

​

What Is an IMU (Inertial Measurement Unit)?

An IMU is a sensor package that combines accelerometers and gyroscopes to measure acceleration and angular velocity along three axes. Commonly installed in robots, drones, vehicles, and other mobile platforms, an IMU can detect attitude (pitch, roll, and yaw) and motion at a high sampling rate. In other words, it continuously records tilt, orientation, and changes in speed—critical data not only for tracking a platform’s behavior and attitude but also for estimating its position.

The key advantage of an IMU is that it enables self-contained positioning through inertial navigation. By integrating the measured acceleration and angular-rate data over time, the system can compute its current position from the previously known position, velocity, and heading—a process known as dead reckoning or autonomous navigation. For example, when a vehicle temporarily loses GNSS reception, it can rely on onboard gyros and speed sensors to estimate its trajectory until the GPS signal returns.

IMU-based dead reckoning is highly effective over short intervals: from the moment GNSS is lost up to several seconds—or even a few tens of seconds—it can maintain reasonably accurate positioning. However, because sensor errors accumulate (drift) over time, an IMU alone cannot preserve high accuracy indefinitely; its positional precision degrades the longer it operates without external reference.

​

RTK–IMU Integration (GNSS/INS Sensor Fusion)

By combining the absolute positioning power of RTK with the relative positioning power of an IMU, each technology offsets the other’s weaknesses. Data from GNSS (RTK) and the IMU are typically merged with state-estimation algorithms such as a Kalman filter. An advanced Kalman filter fuses these measurements to create a GNSS/INS integrated navigation system that estimates position, velocity, and attitude with higher accuracy and stability than either sensor can provide on its own.

In practice, the system interpolates the absolute positions obtained from GNSS with the relative motion derived from the IMU, while any drift accumulating in the IMU is periodically corrected with fresh GNSS updates.

Another hallmark of the fusion algorithm is its dynamic mode-switching based on satellite-signal availability. During normal operation, high-precision RTK fixes are the primary source, with the IMU supplying auxiliary data. If GNSS corrections or fixes are lost, the solution instantly falls back to IMU-based dead reckoning and maintains positioning until GNSS service resumes. Once the signal returns, GNSS data are reintegrated, and IMU errors are re-calibrated.

This architecture enables uninterrupted positioning even during short GNSS outages. Indeed, today’s high-performance receivers already include “hold” functions that keep centimetre-level accuracy for several seconds without corrections, and IMU integration that bridges momentary communication delays.

​

Example Applications of RTK + IMU Fusion

Positioning that blends RTK with an IMU excels in environments where stand-alone GNSS struggles. Below are some typical use cases.

In urban settings—such as beneath elevated roadways or inside tunnels—standard GNSS positioning is typically unavailable. By fusing RTK with IMU data, however, you can estimate the vehicle’s path seamlessly throughout these gaps. In the figure above, the left panel illustrates a tunnel scenario: GNSS fixes (green dots) are obtained at the entrance; inside the tunnel, where no satellite signals are received, the vehicle’s position is continuously calculated by IMU dead-reckoning (yellow dots); at the exit, the system switches back to GNSS. The right panel shows operation in a downtown “urban canyon,” where stable positioning is maintained even under satellite shadowing and multipath effects from tall buildings, thanks to the combined use of GNSS and dead reckoning.

​

Applications in Tunnel Construction and Underground Surveying

At tunnel-boring sites and in underground mapping, positioning has traditionally been difficult because GNSS signals are unavailable. By employing RTK + IMU fusion technology, however, you can maintain an uninterrupted position estimate even deep inside a tunnel.

For instance, in tunnel construction, once the boring machine leaves a known surface control point and enters the tunnel, IMU-based dead reckoning can still track its advance distance and heading with high accuracy. The high-precision start point obtained via RTK at the surface serves as a reference; underground, the IMU continually integrates relative motion to update the position. When the machine reaches a location where GNSS becomes available again—such as the tunnel exit or a ventilation shaft—the accumulated error is corrected, maintaining overall positioning accuracy.

Likewise, in underground surveying, using a GNSS receiver equipped with an IMU enables continuous positioning from the surface down into subterranean spaces, allowing survey work—formerly interrupted whenever GNSS coverage was lost—to proceed smoothly.

​

Use Cases in Skyscraper Districts and Bridge Inspections

In downtown “urban canyons” hemmed in by tall buildings, the limited sky view can make positioning unstable even with multi-GNSS. Here, RTK + IMU fusion shows its strength. A receiver that blends GNSS with an IMU keeps an inertial position solution running in parallel while satellite signals are available; if the satellite count falls or an RTK fix is momentarily lost, it immediately switches to dead-reckoning, ensuring continuous and accurate positioning.

For instance, during infrastructure inspections conducted while walking beneath elevated roadways or in the shadow of buildings, the system maintains uninterrupted, centimeter-level accuracy, enabling precise logging of survey points. In bridge inspections, GNSS reception disappears under the deck or inside girders, but the IMU continues to estimate position using the last GNSS fix as a reference. This lets inspectors map their route accurately and later pinpoint any anomalies they discover.

​

Applications to Mobile Platforms (Vehicles, Drones, Robots)

RTK + IMU fusion is widely used for high-precision navigation of moving platforms. For example, equipping a drone with RTK-GNSS and an IMU allows real-time, centimetre-level self-positioning during flight, so aerial survey and mapping missions can capture imagery with far greater positional accuracy. RTK-equipped drones are already being deployed for situational awareness at disaster sites, achieving in-flight centimetre-level geolocation that would be difficult with GNSS alone.

In the autonomous-vehicle sector, the combination of RTK and inertial navigation is increasingly being adopted. Self-driving systems need lane-level positional accuracy, so augmentation with RTK—beyond standard GPS—is indispensable. Recent trials and a few commercial services already transmit correction data over cellular networks to in-vehicle GNSS receivers, enabling moving vehicles to maintain continuous self-positioning with errors of only a few centimetres.

Because an IMU captures a vehicle’s accelerations and turns at a high rate, it can maintain lane-level accuracy for tens of seconds even when the vehicle passes through tunnels or behind tall buildings; once GNSS is reacquired at the tunnel exit, the accumulated error is corrected instantly.

RTK + IMU fusion also underpins high-precision navigation for outdoor work robots and automated guided vehicles (AGVs). In agriculture, for example, field robots use multi-GNSS RTK to establish a centimetre-level reference position, while the IMU compensates for ground slope and vibration to keep the robot on a straight autonomous path. On construction sites, autonomous machinery relies on the IMU to hold its position when GNSS reception is weak, ensuring uninterrupted accuracy during operations.

In short, blending RTK with IMU technology dramatically boosts the reliability of positioning and motion control for mobile platforms—on land or in the air.

​

LRTK — Key Benefits and Field Advantages

To make RTK + IMU fusion technology easy to deploy on site, Lefixea — a Tokyo Institute of Technology spin-off — developed the LRTK series. LRTK is a pocket-sized RTK-GNSS receiver that pairs with a smartphone or tablet, letting every field worker achieve centimetre-level accuracy as needed. Its main features and the benefits they deliver are outlined below.

• Easy, high-precision positioning via smartphone integration:
  LRTK clips onto the back of a smartphone and connects over Bluetooth or Wi-Fi. Simply launch the dedicated app and tap a button to start positioning. Because the familiar phone interface is used, the device is intuitive even for non-survey professionals, enabling a “one worker, one unit” model in which anyone on site can obtain centimetre-level coordinates instantly whenever needed.

• Compact, lightweight, all-in-one convenience:
  The antenna, battery, and receiver are integrated in a single housing that weighs only about 125 g and is roughly 13 mm thick. Small enough to slip into a pocket, it can be carried in one hand alongside a smartphone while you walk around taking measurements. With no external accessories to lug around, it stays out of the way and is ideally designed for constant, high-precision positioning in the field.

• Multi-GNSS, multi-frequency capability:
  Unlike low-cost single-band receivers, LRTK is a full-featured unit that tracks several constellations—U.S. GPS, Japan’s QZSS (Michibiki), Russia’s GLONASS, Europe’s Galileo, and China’s BeiDou—across multiple frequency bands. Simultaneous reception from many satellites means that if signals are blocked in one direction, others can fill the gap, allowing RTK fixed solutions to be acquired quickly and remain stable. Even in partially obstructed settings—such as beneath tree cover or in the shadow of buildings—LRTK’s multi-GNSS support greatly increases the chance of maintaining high-precision positioning.

• IMU-enabled continuity (dead reckoning):
  Each LRTK model integrates a MEMS IMU that offers a hold function, keeping centimetre-level accuracy for several—up to a dozen—seconds whenever GNSS corrections drop out. Thus, even brief passages through tunnels or under overpasses yield smooth, jump-free positioning. The IMU also provides tilt-compensation: when the receiver is mounted on a survey pole, it automatically projects the pole-tip coordinates straight down, even if the unit is tilted. This lets you work in tight spots or with a slightly angled device without sacrificing accuracy.

• Positioning outside cellular coverage (satellite-based augmentation):
  Standard RTK requires correction data from a reference station delivered over the internet, but mobile service can vanish in mountains or underground sites. LRTK addresses this by supporting an optional antenna that receives Japan’s Quasi-Zenith Satellite System (QZSS) centimetre-level augmentation service (CLAS signal) directly from orbit. Even in remote mountain terrain or tunnel construction zones where a phone has no signal, the receiver can keep RTK positioning alive with satellite-borne corrections. In places where high-precision work once had to be abandoned for lack of communications infrastructure, LRTK now provides stable, centimetre-level accuracy regardless of location.