# A methodological framework for validating a multi-domain physiological sensor for divers using a scalable data fusion platform

## TL;DR

This paper introduces a method to test a new physiological sensor for divers in different environments using a scalable data platform.

## Contribution

A scalable data fusion framework for validating multi-domain physiological sensors in underwater and hyperbaric conditions.

## Key findings

- Human subject testing was conducted in three environments: dry normobaric, dry hyperbaric, and shallow water immersion.
- A custom ROS 2-based data acquisition platform enabled synchronized collection of heterogeneous physiological data.
- The methodology provides a reproducible approach for evaluating sensor performance in complex underwater conditions.

## Abstract

Divers face numerous physiological hazards that can lead to diving disorders, yet the underlying mechanisms remain poorly understood due to limitations in monitoring technologies capable of functioning underwater. To bridge this gap, it is necessary to critically evaluate novel monitoring technologies and develop approaches to determine their suitability for near real-time physiological assessment. Gold standard devices are typically used to validate new physiological sensors under controlled normobaric conditions, where their accuracy benefits from stable, low-interference environments. However, the absence of underwater gold standards and the complexity of underwater conditions make it difficult to determine whether measurement discrepancies arise from true physiological changes or sensor variability. Given the wide range of physiological responses elicited by diving, it is essential to deconstruct the underwater experience into its constituent factors to better isolate their individual effects. This article presents a comprehensive methodology for evaluating the performance of a novel physiological sensor designed for both land-based and underwater environments. To this end, human subject testing was conducted across three experimental environments: 1. dry normobaric testing, 2. dry hyperbaric testing, and 3. shallow water immersion. To facilitate these evaluations, a custom data acquisition platform was developed on Robot Operating System 2 (ROS 2), enabling coordinated synchronization of multiple heterogeneous data streams. This approach offers a scalable and reproducible framework for validating physiological monitoring technologies.