An Artificial Intelligence based Optical Sensor for Microplastic Detection in Seawater

The presence of microplastics (MPs) in marine water is one of the main contamination sources nowadays, becoming a significant threat to marine ecosystems. MPs detection and removal methods are essential to control water quality and implement measures to ensure the preservation of the environment. This paper proposes an optical MP sensor to identify these particles, differentiating them from other elements naturally present in seawater. It employs both an optical detection system and an image acquisition system to detect particles and take several consecutive frames. Subsequently, the sensor's software applies Artificial Intelligence (AI) based models and algorithms to analyze these frames and identify the particle as MP or a different kind of object. Two suitcases house the entire device to ensure portability, IP67 protection against hazardous environmental phenomena for deployments in real locations (corrosion, humidity, etc.), and long-term operation without continuous supervision or maintenance. The design is based on commercial components to create an affordable sensor but capable of analyzing a permanent water flow and uploading results to a cloud database with low latency. The conducted validation reports a promising efficiency for the identification of MPs, which must be confirmed after the final deployment in a real location.