The internal quality assessment of potato tubers is a crucial task in agro-laboratory processing. Traditional methods struggle to detect internal defects such as hollow heart, internal bruises, and insect galleries using only surface features. We present a novel, fully modular hybrid AI architecture designed for defect detection using RGB images of potato slices, suitable for integration in laboratory. Our pipeline combines high-recall multi-threshold YOLO detection, contextual patch validation using ResNet, precise segmentation via the Segment Anything Model (SAM), and skin-contact analysis using VGG16 with a Random Forest classifier. Experimental results on a labeled dataset of over 6000 annotated instances show a recall above 95% and precision near 97.2% for most defect classes. The approach offers both robustness and interpretability, outperforming previous methods that rely on costly hyperspectral or MRI techniques. This system is scalable, explainable, and compatible with existing 2D imaging hardware.

The maritime industry, essential for international logistics and trade, increasingly requires precise and reliable control systems for surface vessels, especially during critical operations such as docking and maneuvering in tight areas. This research presents a data-driven framework for predicting ship control commands, specifically focusing on the number of turns and rudder angle, by using hybrid models that combine deep learning and ensemble machine learning techniques. CNNs are used for feature extraction, while recurrent architectures like GRU and LSTM are used to handle the time-series data generated by a three-DOF vessel model operating in a simulated port environment. To improve prediction performance, these deep learning models are combined with Random Forest and Extra Trees regressors. Comparative evaluations show that the GRU+ExtraTree model delivers superior accuracy and responsiveness, particularly during sudden changes in control signals, due to its ability to capture temporal dependencies. The proposed method shows promising potential for real-time trajectory tracking and autonomous docking applications in maritime navigation.