⚡ electrical engineering

Inter-System Agreement and Reliability of a Gnss-Rtk and Dual-Imu Wearable System for Soccer Performance Monitoring

This study demonstrates that a GNSS-RTK and dual-IMU wearable system offers acceptable inter-system agreement and good-to-excellent interunit reliability for monitoring key soccer external-load metrics like total distance and peak speed, though metrics for very high-speed running exhibit greater variability and require cautious interpretation.

Original authors: Saeid Rostami¹, Mohammad Reza Yazdchi², Bijan Rajaeian³, Amir Shirazi⁴, Mohammad Ali Hashemipoor⁴, Iman Kamali

Published 2026-08-14

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Saeid Rostami¹, Mohammad Reza Yazdchi², Bijan Rajaeian³, Amir Shirazi⁴, Mohammad Ali Hashemipoor⁴, Iman Kamali

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are a coach trying to keep your soccer team in peak shape. You need to know exactly how far they run, how fast they sprint, and how hard they push themselves during a game. In the past, coaches had to guess or rely on simple stopwatches, but today, we have "wearable tech"—little gadgets athletes strap on that act like personal spies, tracking every step. These gadgets use something called GNSS (Global Navigation Satellite System), which is basically the same technology that tells your phone where you are. However, regular GPS can sometimes get a little fuzzy when someone is sprinting or changing direction quickly, kind of like how a map might glitch if you're driving too fast. To fix this, scientists are testing a new, super-smart version of these gadgets that uses "RTK" (Real-Time Kinematic) technology. Think of RTK as a GPS that has a direct, high-speed phone line to a fixed tower nearby, allowing it to correct its position instantly, while also using tiny motion sensors (IMUs) to feel the movement even when the satellite signal gets shaky. The big question is: Do these high-tech gadgets actually agree with each other? If two different coaches wear different trackers, will they get the same numbers, or will one think the player ran a mile and the other think they ran a mile and a half?

This study set out to find the answer by putting a brand-new, fancy "GNSS-RTK and Dual-IMU" wearable system to the test against a well-known, standard 10-Hz GPS device. They asked thirty professional male soccer players to wear both devices at the same time while they did everything from warm-up drills to full-blown competitive matches. The researchers wanted to see if the new gadget could tell the same story as the old one, especially when the players were running at top speed.

The results were a mix of "great news" and "proceed with caution." When it came to the big picture numbers—like the total distance a player ran during the whole game or their absolute fastest speed—the two gadgets were best friends. They agreed almost perfectly. The new system showed a tiny difference of just -6 meters in total distance and +0.08 km/h in peak speed compared to the standard device. If you look at the correlation, which measures how closely their numbers matched, they were very tight: 0.89 for total distance and 0.84 for peak speed. This means that for general workload tracking, the new system is a reliable tool that coaches can trust.

However, the story gets a bit wobbly when the players start sprinting like crazy. When the researchers looked at "Very High-Speed Running" (VHSR)—those short, explosive bursts of speed—the two gadgets started to disagree more. The correlation dropped to 0.65, which is still okay, but not as strong as the other metrics. The difference in how much distance they recorded for these sprints could swing wildly, with a gap ranging from -56 meters to +75 meters. It's like two friends watching a race; they both agree on who won and how long the race took, but when it comes to counting exactly how many steps the winner took in the final sprint, they start to guess differently. The study found that the new system is very reliable (meaning if you use two of them, they give similar results), but the "noise" or error gets bigger the faster the player runs.

So, what does this mean for the soccer world? The authors suggest that this new wearable system is a solid choice for monitoring how much players move overall and how fast they go on average. It's great for tracking the "cumulative" load, like the total miles run in a match. But, when it comes to those split-second, high-intensity sprints, coaches should be careful. They shouldn't rely on a single number from the gadget to make big decisions about a player's health or return to play. Instead, they should look at the trends over time and maybe combine the data with other observations. The paper doesn't claim this gadget is perfect or that it can replace a human eye for every single sprint, but it does prove that for the heavy lifting of tracking soccer performance, this new tech is ready for the field, as long as we treat the sprinting stats with a little extra skepticism.

Technical Summary: Inter-System Agreement and Reliability of a GNSS-RTK and Dual-IMU Wearable System for Soccer Performance Monitoring

Problem Statement
Global Navigation Satellite System (GNSS) wearable technologies are standard in team sports for monitoring external load (e.g., total distance, peak speed, speed zones). However, traditional GNSS devices (typically 1–10 Hz) often exhibit reduced accuracy during rapid accelerations, decelerations, and high-speed running due to positional errors and signal noise. While Real-Time Kinematic (RTK) GNSS technology and Inertial Measurement Units (IMUs) have been integrated to improve positioning and robustness, there is limited evidence regarding the inter-system agreement and interunit reliability of these advanced systems under ecologically valid, soccer-specific conditions. Most validation studies rely on laboratory-based motion capture (a "gold standard"), which does not reflect the practical reality where coaches compare outputs from different commercial field-based systems. This study addresses the gap in understanding how a GNSS-RTK/dual-IMU system agrees with a standard commercial 10-Hz GNSS device during actual soccer activities.

Methodology

Design: A repeated-measures study conducted in outdoor, open-sky environments to simulate applied team-sport monitoring.
Participants: Thirty well-trained male team-sport athletes (professional/semi-professional level, mean age 25.6 ± 3.4 years).
Instrumentation:
- Test System: A wearable device featuring a multi-band GNSS receiver in RTK mode (10 Hz raw data) integrated with dual IMUs (tri-axial accelerometers, gyroscopes, and magnetometers at 1000 Hz). The system utilized sensor-fusion algorithms to output navigation solutions at 25 Hz. RTK corrections were derived from a base station <5 km away.
- Reference System: A commercially available 10-Hz GNSS device (Catapult Vector), widely used in team sport research, served as the comparison standard.
- Internal Load: Polar H10 heart rate monitors were used for exploratory correlation analysis.
Protocol: Athletes performed sport-specific drills, simulated match activities, and competitive match-play scenarios.
Variables: External-load metrics included total distance, peak speed, and distance covered in Low-Speed Running (LSR <14 km·h⁻¹), High-Speed Running (HSR 14–19.8 km·h⁻¹), and Very High-Speed Running (VHSR >19.8 km·h⁻¹).
Statistical Analysis:
- Inter-system Agreement: Assessed via paired-samples t-tests, Pearson correlation coefficients, and Bland-Altman analysis (mean bias and 95% limits of agreement).
- Interunit Reliability: Evaluated using Intraclass Correlation Coefficients (ICC [2,1]) and Percentage Typical Error of Measurement (%TEM).

Key Results

Inter-System Agreement:
- No significant differences were found between the GNSS-RTK system and the reference device for any external-load variable ( $p > 0.05$ ).
- Strong Agreement: Observed for cumulative metrics: Total Distance ( $r = 0.89$ ), Peak Speed ( $r = 0.84$ ), and HSR Distance ( $r = 0.81$ ). Bland-Altman analysis showed small mean biases for these variables (e.g., Total Distance bias: −6 m).
- Weaker Agreement: VHSR Distance demonstrated lower correlation ( $r = 0.65$ ) and wider limits of agreement (Mean bias: +9 m; 95% LoA: −56 to +75 m), indicating greater variability at high intensities.
Interunit Reliability:
- The GNSS-RTK system showed good-to-excellent reliability (ICC = 0.80–0.96) across most variables.
- Measurement Error: %TEM was lowest for time-based and low-speed variables (2.7%–5.7%) but significantly higher for VHSR metrics (11.2%–13.2%).
- The reference device exhibited similar error patterns, with VHSR metrics showing the highest variability in both systems.
External-Internal Load Relationships:
- Total distance, peak speed, and HSR distance showed moderate-to-strong correlations with heart rate ( $r = 0.55–0.71$ ).
- VHSR distance showed non-significant correlations with heart rate ( $r = 0.34–0.37$ ), attributed to the physiological lag of heart rate during short, explosive movements.

Key Contributions

Field-Based Validation: Provides empirical data on the agreement between a high-frequency GNSS-RTK/dual-IMU system and a standard commercial GNSS device in real-world soccer conditions, rather than relying on laboratory motion capture.
Differentiation of Metrics: Clearly delineates that while cumulative and moderate-intensity metrics (Total Distance, Peak Speed, HSR) are highly reliable and agreeable, VHSR metrics suffer from substantial measurement variability and lower agreement.
Reliability Benchmarking: Establishes that the observed variability in high-speed running is not unique to the new RTK system but appears to be an inherent limitation of GNSS-based tracking during rapid accelerations, as the reference device showed similar error patterns.

Significance and Claims
The paper concludes that the GNSS-RTK and dual-IMU wearable system is suitable for practical application in monitoring cumulative external load and moderate-intensity running in soccer. The system demonstrates acceptable inter-system agreement and good-to-excellent interunit reliability for total distance and peak speed.

However, the authors explicitly state that caution is required when interpreting Very High-Speed Running (VHSR) metrics. Due to the reduced agreement and increased measurement variability at high velocities, these specific metrics should not be over-relied upon for individual decision-making or acute load management without complementary data. The study emphasizes that its findings represent inter-system agreement rather than absolute validity, as no gold-standard criterion system (e.g., optical motion capture) was used. The results support the use of this technology for general workload monitoring but highlight the need for careful interpretation of high-intensity sprint data.

Technical Summary: Inter-System Agreement and Reliability of a GNSS-RTK and Dual-IMU Wearable System for Soccer Performance Monitoring

More like this