EasySort

About the Project

EasySort is an AI-powered waste sorting Android application developed by Team NEXUS², consisting of two secondary school students from Penang, Malaysia. The project aims to improve recycling awareness and help users identify waste materials through a smartphone camera. EasySort uses a custom-trained YOLO11 image classification model developed using our own waste image dataset. The newer model contains 10 trained classes: Aluminium, Background, Battery, Ceramic Cup, Crushed Paper, Paper, Paper Cup, Phone, Plastic Bottle and Plastic Cup. Nine classes represent waste items, while Background represents a non-waste class. The application provides waste recognition for supported items, recycling guidance, educational tips and information about recycling locations. It was developed using Kotlin, Jetpack Compose and TensorFlow Lite to support on-device classification. Our team is also exploring the integration of pretrained model capabilities to improve recognition reliability. Further testing is required to measure any accuracy improvements. A user survey involving at least 50 respondents found that approximately 95% considered EasySort easy to use. This result reflects user feedback rather than measured classification accuracy. EasySort aims to encourage responsible waste disposal and promote environmental awareness. The project supports United Nations Sustainable Development Goal 12: Responsible Consumption and Production, and the Penang Green Agenda 2030.