Deep Learning-Enabled Dissolved Oxygen Sensing in Biofouling Environments for Ocean Monitoring

2026-04-27

AI Summary

This study addresses the challenge of signal drift and limited long-term accuracy in low-cost dissolved oxygen sensors operating in marine biofouling environments. To overcome this, the authors propose a novel approach that integrates camera-based optical sensing with physics-informed neural networks. By embedding the Stern–Volmer equation into the loss function of a Vision Transformer (ViT) and employing a deep ensemble strategy, the method achieves, for the first time, physically constrained dissolved oxygen measurements robust to biofouling. The approach substantially improves measurement accuracy—reducing mean absolute error by 89–92% compared to conventional statistical and machine learning methods, achieving approximately 2 μmol/L—and further enables self-diagnosis and quantification of predictive uncertainty.

Technology Category

• Computer Vision: Low Level & Physics-based Vision
• Intelligent Robots: State Estimation
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Security and Privacy: Data transparency and provenance

Abstract

The escalating climate crisis and ecosystem degradation demand intelligent, low-cost sensors capable of robust, long-term monitoring in real-world environments. Absolute dissolved oxygen (DO) concentration is a key parameter for predicting climate tipping points. Inexpensive optoelectronic sensors based on microstructured polymer films doped with phosphorescent dyes could be readily deployable; however, signal drift and marine biofouling remain major challenges. Here, we introduce a sensing paradigm that combines camera-based DO sensors with a visual transformer (ViT)-based physics-informed neural network (PINN) for high-fidelity sensing under biofouling conditions. Training and testing data were obtained from an algae-laden water tank over 14 days to capture accelerated biofouling. The ViT-PINN, which embeds the Stern-Volmer (SV) equation into the loss function, reduces mean average error (MAE) by 92% and 89% compared to classical statistical and ML approaches, achieving ~2 umol/L absolute error. A deep ensemble further quantifies predictive uncertainty, enabling self-diagnostic sensing.

Problem

• dissolved oxygen sensing
• biofouling
• ocean monitoring
• signal drift
• optoelectronic sensors

Innovation

• physics-informed neural network
• visual transformer
• biofouling-resistant sensing
• dissolved oxygen monitoring
• Stern-Volmer equation