We can't find the internet
Attempting to reconnect
Something went wrong!
Hang in there while we get back on track
Detecting Microplastics in Seawater with a Novel Optical Sensor Based on Artificial Intelligence Models
AI summary Read the abstract
Detecting microplastics in seawater quickly and accurately is a major technical challenge, and this study developed a novel optical sensor that uses artificial intelligence to identify plastic particles from light-scattering data in real time. The AI-powered system was tested on seawater samples and showed promising accuracy for classifying microplastic types without the need for time-consuming laboratory processing. Automated in-situ sensors like this could enable continuous, large-scale ocean monitoring for microplastic pollution.
Ocean pollution is a significant and growing menace, not only to marine environments, but to the entire planet due to the interconnected nature of biosphere ecosystems. Every day, vast amounts of plastic waste are dumped into the oceans and, more specifically, millions of microplastics endanger marine flora and fauna in the short, medium, and long term. Finding solutions to detect, characterize, and remove these tiny particles is a critical challenge that requires continuous innovation in the coming years to protect seas and oceans. In this context, this paper presents a microplastic detection and identification sensor based on optical principles for particle detection, combined with microscopy and artificial intelligence algorithms for identification, differentiating microplastics from other natural or synthetic particles. An innovative approach enables the inspection of continuous water flows and, while originally designed as a benchtop sensor, minor adaptations can leverage this feature to create a portable version suitable for deployment in real-world scenarios. The sensor requires comprehensive training of detection algorithms, followed by validation tests to optimize efficiency in each deployment, whether at the laboratory scale or in real-world scenarios. The objective is to achieve efficiencies greater than 85–90% in both cases. To this end, two deployments have been carried out (one of which is still ongoing), with the aim of extracting lessons to optimize the sensor and ultimately develop a final version that can aid in microplastic detection and the preservation of marine ecosystems.
More Papers Like This
An Artificial Intelligence based Optical Sensor for Microplastic Detection in Seawater
AI summary Read the abstract
Researchers developed an AI-based optical sensor system combining an optical detection subsystem and an image acquisition subsystem to detect and identify microplastic particles in seawater, distinguishing them from naturally occurring marine particles. The device applies AI algorithms to analyze consecutive image frames and classify particles as microplastic or non-microplastic, with the full system housed in two portable cases.
Real-time detection of microplastics in aquatic environments using emerging technologies
AI summary Read the abstract
Researchers proposed a real-time microplastic detection system combining AI-enhanced optical sensors and IoT devices, capable of automatically classifying microplastics in ocean water without the time-consuming manual steps required by spectroscopy or microscopy.
The Development of Sensors for Microplastic Detection Using Artificial Intelligence
AI summary Read the abstract
This review examined AI-enhanced sensors developed for microplastic detection and characterization in aquatic environments, covering machine learning, deep learning, and spectroscopic sensor approaches. The authors found that AI substantially reduces the labor intensity of microplastic identification and improves detection of small particles, though training dataset standardization and real-world validation remain priority challenges.
Automatic Identification and Classification of Marine Microplastic Pollution Based on Deep Learning and Spectral Imaging Technology
AI summary Read the abstract
Researchers developed an AI system combining deep learning with multispectral imaging to automatically identify and classify marine microplastics, using a feature-selection method called ReliefF to reduce noise in complex ocean samples. The approach achieved high accuracy and offers a scalable solution for large-scale ocean microplastic monitoring that outperforms traditional manual inspection.
Toward in Situ Identification of Microplastics in Water Using Raman Spectroscopy and Machine Learning
AI summary Read the abstract
This study developed an early-stage system combining Raman spectroscopy and machine learning to identify microplastics directly in ocean water in real time, without needing to collect and process samples in a lab. A support vector machine classifier trained on spectral libraries correctly identified all pristine microplastic samples and most environmental ones, demonstrating that field-deployable automated detection is feasible. Accurate real-time monitoring tools are urgently needed to understand where microplastics concentrate in the ocean and to track pollution trends.
Research digests by email
When a large batch of papers lands in the Atlas, we read through it and send a short write-up of what stood out.