Abstract:
This research develops a Smart Waste Management System based on the Internet of Things (IoT) with Convolutional Neural Network (CNN) integration to improve the efficiency of waste sorting and management. The system is designed using the ESP32-CAM module to capture images of waste, then sends it to a local Flask based server to be classified into four categories: organic, inorganic, medical waste, and hazardous waste (B3). The classification results are used to drive three servo motors that direct waste to the bins according to the category. Four ultrasonic sensors are installed on each keg to detect volume capacity, while indicator LEDs and buzzers provide local alerts when the keg is full. All status data is sent to the Node-RED platform via the MQTT protocol and displayed on the dashboard in real-time. The test results showed that the system was able to classify waste with high accuracy, detect capacity appropriately, and provide responsive notifications. This innovation is expected to be an effective solution in supporting smart, efficient, and sustainable waste management.