<![CDATA[The increasing presence of plastic debris in these areas not only disrupts biodiversity but also threatens the balance and sustainability of marine habitats. Addressing this problem requires innovative approaches that combine environmental science with modern technology. This study was conducted to develop a beach litter detection system as part of a broader effort to support environmental preservation and reduce the detrimental effects of plastic waste on coastal areas in Indonesia. The research employed a secondary dataset obtained from Kaggle.com, which consisted of labeled images of beach waste. A deep learning method was applied through the use of Convolutional Neural Networks (CNN), with MobileNetV2 selected as the primary architecture due to its lightweight design, computational efficiency, and proven effectiveness in image classification tasks. Experimental results demonstrated that the model performed exceptionally well, achieving a training accuracy of 100%, which indicates its strong ability to capture patterns in the dataset. More importantly, the validation accuracy reached 97.83%, reflecting the model’s robustness and capacity to generalize effectively to unseen data. These findings emphasize the potential of artificial intelligence in supporting environmental monitoring and management. In particular, automated detection and classification of plastic waste on beaches can enhance current conservation strategies and provide timely information for waste management interventions. Furthermore, this research serves as a foundation for future studies aimed at advancing intelligent waste management systems. The integration of AI in this domain remains relatively underexplored, and continued exploration could contribute significantly to mitigating the global challenge of plastic pollution in coastal environments.]]>
