Towards an Autonomous Surface Vehicle Prototype for Artificial Intelligence Applications of Water Quality Monitoring

TL;DR

This paper presents a prototype autonomous surface vehicle equipped with sensors and AI for water quality monitoring and macro-plastic detection, demonstrating real-world effectiveness in environmental water resource applications.

Contribution

It introduces a fully equipped vehicle prototype integrating advanced sensors and deep visual models for water quality and macro-plastic detection in real environments.

Findings

Successful detection of macro-plastics using deep visual models

Effective measurement of water quality parameters

Prototype demonstrated in real environmental conditions

Abstract

The use of Autonomous Surface Vehicles, equipped with water quality sensors and artificial vision systems, allows for a smart and adaptive deployment in water resources environmental monitoring. This paper presents a real implementation of a vehicle prototype that to address the use of Artificial Intelligence algorithms and enhanced sensing techniques for water quality monitoring. The vehicle is fully equipped with high-quality sensors to measure water quality parameters and water depth. Furthermore, by means of a stereo-camera, it also can detect and locate macro-plastics in real environments by means of deep visual models, such as YOLOv5. In this paper, experimental results, carried out in Lago Mayor (Sevilla), has been presented as proof of the capabilities of the proposed architecture. The overall system, and the early results obtained, are expected to provide a solid example of a real platform useful for the water resource monitoring task, and to serve as a real case scenario for deploying Artificial Intelligence algorithms, such as path planning, artificial vision, etc.