Validating Waterproof Wireless IMUs Against Motion Capture: Cometa WaveTrack in Freediving

Studying human movement underwater means working around a hardware constraint that has defined aquatic biomechanics for decades: optoelectronic motion capture, the reference method on dry land, does not transition cleanly into water. Light refraction distorts marker positions, bubbles obstruct the line of sight, and the capture volume rarely covers more than a few square meters of pool. The result is that most underwater studies operate with shorter acquisitions and tighter spatial constraints than the research questions actually require.

Waterproof wireless IMU sensors are the structural alternative. Worn directly on the body, occlusion-free, deployable across full water volumes and sustained acquisition durations, they remove the geometric limits of camera-based systems. The operational question is whether the data they produce is comparable to what an optoelectronic reference would record — because in research-grade work, the wireless system needs to support the same analytical conclusions a camera-based study would. For Cometa WaveTrack IMU sensors operating during freediving propulsion, that question now has a peer-reviewed answer.

A 2026 study published in Sensors and Actuators: A. Physical (Algourdin et al.) compared Cometa wireless IMUs against an underwater Qualisys Arqus A12U system during vertical flutter kicking. On the parameters that drive aquatic propulsion analysis — kicking frequency, sagittal-plane joint amplitude, waveform shape — the two systems converge. The WaveTrack data is research-grade.

Underwater optoelectronic systems can be deployed in pools, but the operational cost is high and the capture window is limited. Refraction at the air-water interface, light attenuation with depth, and persistent occlusion from bubbles and turbulence degrade marker tracking accuracy. For groups studying sustained efforts, multiple subjects, or movements that travel through the water column, the camera layer stops being a viable primary measurement system.

Cometa systems are typically deployed in biomechanics research protocols where wireless EMG and IMU sensors are integrated with optoelectronic motion capture and force plates. On dry land, the camera system handles full-body kinematics and the wireless sensors provide muscle activation and inertial data the cameras cannot capture. The two acquisition layers are complementary.

Underwater, this division of labor shifts. The wireless layer becomes the primary measurement system, and the reliability of that primary layer determines whether the science can be published. The WaveTrack inertial platform was designed for exactly this scenario — sustained acquisition in unrestricted volumes, with the same data integrity as a dry-land setup. Cometa’s broader work on EMG underwater applications has already documented the EMG side of this equation. The Algourdin et al. study goes in depth on the IMU side.

The research team, based at the MAPIEM laboratory of the Université de Toulon set out to characterize the lower-body kinematics of freedivers during vertical flutter kicking and to determine whether Cometa WaveTrack IMUs could carry that work without an optoelectronic reference. The validation was built against an underwater Qualisys Arqus A12U system calibrated to sub-millimeter accuracy.

Seventeen participants of mixed experience, from non-divers to competitive freedivers, performed one minute of vertical flutter kicking in a 7 × 4.5 × 3 m pool. The mixed-experience sample is methodologically intentional: it covers the full range of motor proficiency the sensors are likely to encounter in field deployment.

Each subject was instrumented with twenty-one retro-reflective markers following the Qualisys Sport Marker Set protocol and seven Cometa WaveTrack IMU sensors (43 × 26 × 11 mm, 9.8 g) positioned on the dorsal foot, medial tibia, lateral thigh, and pelvis. Six underwater Qualisys cameras (12 MP, 4096 × 3072 px, 300 fps) operated at 100 Hz.

The methodological detail that matters most is the processing pipeline. Both data streams were run through the same OpenSim model (gait2392_simbody), with inverse kinematics applied to extract joint angles from a single biomechanical reference frame. This eliminates the most common confound in IMU-versus-camera comparisons — different models producing different angles, with the discrepancy attributed to the hardware. What the study compared, in the end, was the sensors themselves.

On kicking frequency, the two systems agree. Relative error stayed below 3% across every joint measured, with both Cometa and Qualisys converging on stroke rates of 0.46 to 0.47 Hz. For studies whose primary outcome is propulsive frequency or energy cost estimation, the wireless and camera-based data are interchangeable.

On joint amplitude in the sagittal plane, the plane of flexion and extension that dominates flutter kicking, the agreement is strong. Mean amplitudes for the ankle, knee, and hip flexion showed relative errors below 30% across the full signal, Pearson coefficients above 0.9, and ICC values up to 0.91 for hip flexion after offset correction. The authors describe the sagittal-plane waveform similarity as excellent.

The frontal and transverse planes — hip rotation, hip adduction, pelvis rotation — show a more cautious reliability profile. Pearson coefficients drop to between 0.71 and 0.85, and relative errors inflate on movements with very small absolute ranges, where minor degree-level differences produce large percentage values.

The honest reading of the data, which the paper itself adopts, is that Cometa wireless IMUs deliver research-grade reliability for the parameters that drive aquatic propulsion analysis — frequency, sagittal-plane amplitude, waveform shape in flexion — and that frontal and transverse plane analysis should be interpreted with awareness of the precision limits documented in the study. This plane-specific reliability profile is consistent with the broader IMU literature for walking and running on dry land. The aquatic environment did not degrade the sensors beyond what is expected of the technology in general.

This research establishes a peer-reviewed reference for subsequent aquatic biomechanics studies, both for groups using Cometa sensors as the primary measurement layer and for those combining them with an underwater Qualisys setup to extend the analytical reach beyond what cameras alone can cover. It complements existing work on stroke cycle analysis using WaveX and IMU sensors, extending the reliability envelope from horizontal swimming to vertical propulsion.

The WaveTrack platform is built on a single acquisition architecture that operates from dry-land sport and clinical contexts to fully immersed aquatic research, with the same data integrity and the same software ecosystem across deployment scenarios. The validation reported here documents one specific case. The methodology is reproducible, the limitations are stated, and the published evidence base for waterproof wireless IMU biomechanics is now one paper denser.