2025_proceedings: Advancing Marine Monitoring Through Edge-Optimised Deep Learning: Hydrotwin’s Approach to Underwater Acoustic Surveillance

Back


  • Session: 08. Bioacoustics and Soundscape
    Organised by: Jennifer Miksis-Olds, Giacomo Giorli
  • Lecture: Advancing Marine Monitoring Through Edge-Optimised Deep Learning: Hydrotwin’s Approach to Underwater Acoustic Surveillance
    Paper ID: 2219
    Author(s): Astrid van Toor, Erica Cruz, Guilherme Beleza Vaz
    Presenter: Astrid van Toor
    Abstract: Advancing sustainable ocean monitoring is critical for preserving marine biodiversity and supporting ecosystem health. The Hydrotwin solution leverages cutting-edge digital twinning technology to integrate underwater sound measurement and modelling with machine learning for real-time marine life monitoring and ecosystem analysis. By coupling environmental parameters with high-resolution underwater acoustic data, the platform provides novel insights into the underwater soundscape. This paper explores the application of Hydrotwin in marine monitoring, emphasizing its potential to bridge gaps in biodiversity assessments through scalable, automated systems. Key contributions include the development of multi-modal data pipelines that combine acoustic signatures, environmental simulations, and species-specific recognition frameworks, powered by deep neural networks for enhanced analysis of marine soundscapes and ecosystem dynamics. Additionally, we highlight the system’s ability to predict environmental impacts, support conservation policies, and enable early detection of anthropogenic disruptions such as illegal fishing activities. Hydrotwin’s integration of predictive modelling, visualisation dashboards, and real-time updates establishes it as a transformative platform for marine science. By enabling dynamic, high-fidelity monitoring solutions, it sets a foundation for actionable insights into marine ecosystems, underscoring the vital role of digital twin technologies in fostering sustainable ocean stewardship.
      Download the full paper
  • Corresponding author: Ms Astrid van Toor
    Affiliation: BlueOasis
    Country: Portugal

Back