Boat wakes introduce a significant amount of energy into the littoral zone of a lake ecosystem. The impact of wake energy on the shoreline leads to a deterioration of water quality due to sediment resuspension. A noticeable decline in macrophyte vegetation has already been observed at Lake Wörthersee, Carinthia. Within the interdisciplinary research project WAMOS (Wave Monitoring System based on GNSS/INS Integration) the impact of boat wakes on the aquatic vegetation and, consequently, on the water quality of Lake Wörthersee was investigated.

Within the paper a wave-measuring buoy system based on low-cost inertial sensors is presented to determine the heights of boat-induced waves. The measurement system consists of buoys equipped with inertial measurement units (IMU) capable of capturing both sensor orientation and vertical displacement. A modified strapdown algorithm incorporating a rotation-angle-based attitude Kalman filter is used and described. Using trend correction steps driven by regression estimation, the algorithm is able to determine wave heights from the integration of the inertial measurement data with appropriate accuracy in the time domain. To verify the wave height determination, the wave heights calculated using the strapdown algorithm were compared with reference data obtained by tracking one of the buoys with a Leica Nova MS60 total station. This comparison demonstrates congruence within the sub-centimeter range for both the water surface elevation and the wave heights derived from it.

The calculated wave height data are utilized to conduct analyses of boat traffic. Time synchronization of measurement data from multiple buoys in a buoy field enables the investigation of wave propagation and wave direction. Additionally, the research encompasses spectral analysis using the wavelet transform, which allows for the visualization of boat-induced waves in the time-frequency domain. Wave signals associated with individual boat passages are clearly separable from ambient noise in the wavelet spectrogram, where their power spectral density forms a distinctive pattern in the time–frequency domain. These characteristics enable conclusions to be drawn about the types of vessels that generated the wakes and their operational mode (displacement or planing). The wavelet spectrograms derived from water surface elevation data obtained through IMU integration were found to agree well with the total station reference measurements, but deviations occur in the low-frequency spectral range due to trend reduction.

This work presents the implementation of a fully operational, autonomous and cost-effective wave measurement system capable of computing boat wakes from inertial data in near real time. The collected measurement data provide a foundation for future environmental protection measures at Lake Wörthersee and contribute to an improved understanding of nearshore wave processes and their impacts on the aquatic ecosystem.

Global Navigation Satellite Systems (GNSS) are assumed to be a cornerstone for navigation of automated vehicles. GNSS correction services are used to improve position accuracy. How do different correction services compare in accuracy and availability? For automated driving GNSS can be supplemented by other sensors like lidars and inertial navigation systems, but how reliable is the GNSS positioning itself?

As part of the EU-funded MODI project, the Norwegian Mapping Authority performed a GNSS receiver data collection campaign while driving the route Oslo–Rotterdam–Oslo in June 2025. Eleven different receivers with seven different correction services were tested and analyzed with respect to various key performance indicators relevant for automated vehicles. Service types tested include Network Real Time Kinematic (NRTK), commercial and national Precise Point Positioning RTK (PPP-RTK) and Galileo High Accuracy Service (HAS).

The well-proven NRTK services show availability of precise positions for 92% of epochs, with 95% of converged epochs having horizontal position errors less than 4cm (HPE95 of 4cm). Commercial PPP-RTK services varied in performance, with availability of 91% and 57% and HPE95 of 12cm and 10cm. National PPP-RTK achieved 83% availability with HPE95 of 13 cm. Galileo HAS performed comparably to PPP-RTK in static tests but showed poor kinematic performance (2-76% availability depending on receiver), likely related to long convergence times.

Service availability – the percentage of epochs where the receiver reports its most precise positioning mode – is critical for automated driving. The study investigated durations and causes of service gaps, finding that bridges, tunnels, and gas station stops were common causes. With availability of 89-92% for the key correction services, alternative positioning methods are clearly needed to fill the gaps.

The duration of anomalies (periods of large position errors) is another important consideration. Results show that most anomalies for NRTK and PPP-RTK services last only five seconds or less, though longer anomalies occasionally occur for all services.

Additional challenges were identified, including unique national realizations of the ETRS89 reference frame for every country, highlighting the need for standardization and documentation of reference frames and their EPSG codes.

In summary, achieving stable and accurate GNSS-based positioning for automated driving over long distances and across borders remains challenging. Key issues include variations in positioning service types, their availability and accuracy, and differences in geodetic
reference frames.

In recent years, high-precision Global Navigation Satellite System (GNSS) positioning using low-cost portable devices has attracted significant research interest. Specifically, the release of Android 7 (Nougat) in 2016 by Google made it possible to obtain typical GNSS parameters directly from raw measurements, bringing significant innovation in smartphone location services.

While GNSS-based positioning on smartphones has been widely investigated, the growing demand for accurate location services in sports and health-related applications has shifted increasing attention toward GNSS-enabled smartwatches. Raw GNSS measurements can now be logged on Android-based smartwatches. Despite this capability, the data currently suffers from several issues, such as discontinuities in signal tracking and increased noise in observations. Nevertheless, only a limited number of studies have conducted systematic assessments of raw measurement quality on smartwatches. Within the existing literature, smartphone-oriented assessment approaches are commonly adopted, focusing on the signal strength, and providing general analyses of pseudorange and carrier-phase noise. Crucially, they did not thoroughly evaluate inconsistencies among measurements, systematically performe cycle-slip detection for carrier-phase observations, or separately examine the impact of multipath effects on the measurements.

Addressing these gaps is critical because raw data collected from smartwatch is fraught with uncertainties. Specifically, the Android API does not provide pseudorange measurements directly, and different conversion tools lead to observation discrepancies. Moreover, the compact design of smartwatches imposes severe constraints on antenna size and placement, potentially increasing their sensitivity to multipath effects. Furthermore, unstable signal tracking due to inherent limitations leads to potential cycle-slips between consecutive carrier-phase measurements.

To bridge these gaps and support optimized GNSS data processing on smartwatch, this study introduces a comprehensive framework for assessing the quality of raw smartwatch GNSS measurements. We evaluate observations from the Samsung SM-R940 smartwatch and compare them with those from the Pixel 7 Pro smartphone, which serves as a reference. Specifically, the assessment focuses on the following aspects:

1. Carrier-to-noise density ratio (C/N0):
C/N0 is utilized to characterize the received signal strength. We evaluate its magnitude and continuity for each satellite, and analyze its dependence on satellite elevation.

2. Measurement continuity:
We assess the availability of pseudorange, carrier-phase, and Doppler measurements for each satellite, and quantify the percentage of discontinuities.

3. Inconsistency among measurements:
Measurement inconsistencies are analyzed using first-order differences of pseudorange and carrier-phase observations, alongside Doppler-derived ranges. When raw log files are used as input, pseudorange, carrier-phase, and Doppler measurements are generated using the algorithm provided by the European GNSS Agency.

4. Multipath effects:
Geometry-free combinations (pseudorange minus carrier-phase measurements) and carrier-phase-smoothed pseudorange errors are used to detect significant multipath effects. Epochs exhibiting rapid fluctuations in these metrics are identified as being strongly affected by multipath effects.

5. Cycle-slip detection:
Cycle-slips are detected by analyzing the first- and third-order Time-Differenced Carrier-Phase (TDCP) measurements, alongside carrier-phase predicted errors. In the absence of cycle-slips, these errors remain stable; when a cycle-slip occurs, both the TDCP series and the predicted error show pronounced jumps.

Overall, this study lays the foundation for developing robust preprocessing strategies and high-precision positioning algorithms tailored for smartwatches.