Proyecto de investigación

Monitorization of Environmental Dangers with Unmanned Surface Agents: (MEDUSA)

Responsable: Sergio Toral Marín / Daniel Gutiérrez Reina
Tipo de Proyecto/Ayuda: Plan Complementario de I+D+ del PRTR
Referencia: PCM_00019
Fecha de Inicio: 23-01-2023
Fecha de Finalización: 30-09-2025

Empresa/Organismo financiador/es:

  • Junta de Andalucía: Consejería de Universidad, Investigación e Innovación



  • Investigadores:
    • Alejandro Casado Pérez

Resumen del proyecto:

The MEDUSA project proposes a novel monitoring platform for coastal and transitional waters. The monitoring platform consists of a scalable fleet of Autonomous Surface Vehicles (ASVs) equipped with physico-chemical sensors and a cloud centralized system for data storage, visualization, modeling, and forecasting. ASVs are equipped with sensors for Water Quality Parameters (WQP), such as pH, conductivity, ORP (oxygen reduction potential), nitrates, phosphorus, among others, and multispectral cameras to detect persistent organic/biological and plastic waste. The ASVs continuously send the collected data to the global coordinator platform. This Cloud coordinator consists of a server with computational power to: i) process and store data, ii) train and deploy machine and Deep learning models, and iii) visualize data by means of dashboards. Communications among the ASVs and the Cloud coordinator are established by means of 4G or 5G links.

The outputs of the MEDUSA platform can be divided into two categories. On the one hand, those targeted at technical stakeholders; that is, technicians of public administrations, research groups, and the research community in general. Within this category, the MEDUSA project develops:

  • ·        An early alert system that uses the results generated by the online modeling module to detect anomaly values that put the marine ecosystems at risk. This module indicates the locations of the areas where action should be taken to avoid the degradation of marine ecosystems.
  • ·        Forecasting models that predict the evolution of the WQP. These forecasting models will anticipate eutrophication scenarios according to the predicted evolution of the physico-chemical parameters. Autoregressive models will be used for these tasks.
  • ·        Cloud storage of the data collected and the results of the models developed for posterior analysis. Data will be accessible through a user-friendly API.

On the other hand, the outputs of the MEDUSA platform intended for public administration stakeholders are the visualization dashboards that can be easily interpreted by the society and will serve to raise awareness of the status of the coastal areas and the actions that should be taken by public administrators. Within this category, the MEDUSA project develops:

  • ·        A visualization dashboard for the results of the online module, including anomaly detection and color maps.
  • ·        Concentration of solid and surface pollutants to facilitate cleaning operations.
  • ·        A dashboard for the predicted evolution of the WQP, indicating zones at risk of developing eutrophication phenomenon.
Vicerrectorado de Investigación. Universidad de Sevilla. Pabellón de Brasil. Paseo de las Delicias s/n. Sevilla