Abstrak Artikel
Abstract The increasing volume of waste requires efficient sorting approaches to support recycling and sustainable waste management. Manual waste separation is labor-intensive, time-consuming, and susceptible to classification errors. This study develops a microcontroller-based waste sorting system for distinguishing metal and non-metal waste using an inductive proximity sensor as the primary detection device. Two HC-SR04 ultrasonic sensors are integrated to monitor the fill levels of metal and non-metal waste bins, while a servo motor directs the classified waste to the corresponding containers. An LCD module provides real-time information on waste classification and bin capacity. The system was developed using a Research and Development (R&D) approach and evaluated through functional testing, waste classification accuracy testing, and bin capacity monitoring. The experimental results showed that the system achieved a metal waste classification accuracy of 53.8% and a non-metal waste classification accuracy of 100%, resulting in an overall classification accuracy of 73.9%. The results indicate that the inductive proximity sensor provides reliable detection of non-metal waste classification but has limitations in detecting metal-containing objects, particularly those affected by surface covering, material characteristics, and object positioning within the sensor's effective detection range. These findings demonstrate the feasibility of the proposed prototype for small-scale and educational waste sorting applications while highlighting the need for further improvement in sensor placement and mechanical feeding design to improve metal detection reliability.
Pratinjau PDF
Pastikan PyMuPDF terpasang lalu coba muat ulang.