<![CDATA[The lack of attention to electronic waste (e-waste), particularly regarding mobile phones, has a serious impact on global environmental issues. One of the main obstacles in the economic circulation of these devices is the subjectivity and technical difficulty in accurately assessing the physical condition of used phones. This research aims to address these challenges through the development of a circular economy platform prototype based on a mobile application that provides objective and automated phone condition assessment services. The system is designed using React Native Expo and integrates the MobileNetV2 Deep Learning model via TensorFlow Lite. Transfer learning methods are applied to a dataset covering various mobile phone brands such as Samsung, Xiaomi, and OPPO to train the model to recognize physical damage on the screen and body. Test results indicate that the system is capable of providing objective assessment with high precision for devices in prime condition (Grade A) at 0.95. However, objectivity for severely damaged phones (Grade D) remains a challenge with a precision of 0.22 due to training data imbalance. Nevertheless, the application prototype successfully presents a transparent real-time scanning feature. This research contributes to providing a technical solution that bridges the trust gap through automated assessment standardization, thereby minimizing manual inspection subjectivity and promoting supply chain efficiency in the electronic circular economy.]]>
