Chaoyang University of Technology

Due to the deteriorating global environment, waste recycling has become one of the key issues for environmental protection. According to the market research on vending machines by Global Information, Inc. (GII), the market size of vending machines is expected to continue expanding in the future, which also means that the issue of recyclable objects generated from vending machines will become more pressing. In response to this trend, this study focuses on the recycling of the most common products found in vending machines and proposes the development of a system that uses object recognition to assist humans in sorting recyclable materials from vending machines. This not only presents a new business opportunity for vending machine companies but also contributes to environmental protection. The recycling object identification system utilizes YOLOv7-tiny as a training model, using the best-trained weight file with a mean Average Precision (mAP) of 98.9% across all categories. The system is further integrated with HUB8735, which uses camera-based object recognition to control motors for sorting. The operational processes have been modeled and verified using Petri nets, a mathematical and graphical modeling tool, to ensure their accuracy and completeness. Finally, a recyclable vending machine was successfully built and detailed discussion on its operations has been presented. Beyond technical implementation, this research offers a scalable and cost-effective solution for decentralized waste management, demonstrating a significant economic advantage over manual labor. By enhancing the purity of recovered resources and minimizing carbon footprints associated with improper waste disposal, this system provides a practical framework for integrating AI-driven automation into the global circular economy.

Digital clubbing is an important clinical sign associated with a range of cardiopulmonary diseases; however, its detection and severity assessment in routine practice largely rely on subjective visual inspection. This study proposes an automated, smartphone-based system for real-time detection and severity assessment of digital clubbing using deep learning techniques. The system integrates the YOLOv8 object detection model for initial clubbing classification, the KeypointNet model for anatomical landmark localization, and a novel Clubbing Fingers Severity Analysis (CFSA) algorithm to quantify the Lovibond angle and grade disease severity. Finger images were acquired using a smartphone camera with an OpenCV-based preprocessing strategy to standardize finger-to-camera distance and improve image consistency. Model performance was evaluated using publicly available anonymized datasets. The proposed system achieved an overall accuracy of 94.7% for digital clubbing detection and severity classification. The YOLOv8 model attained a classification accuracy of 92.5%, while the KeypointNet model achieved a landmark localization accuracy of 96.5%. Notably, the recall for severe digital clubbing reached 94.0%, indicating strong sensitivity for identifying high-risk cases. By providing real-time, non-invasive, and reproducible assessments, the proposed system addresses the limitations of conventional visual examination and supports objective severity grading. Although further clinical validation is required, this smartphone-based approach demonstrates strong potential as a preliminary screening support tool for early identification of digital clubbing in clinical and community-based settings, particularly in resource-limited environments.