SISTEM MONITORING SAMPAH CERDAS BERBASIS DRONE MENGGUNAKAN VISI KOMPUTER

Abstract

Conventional Municipal Solid Waste (MSW) management, which is reactive and relies heavily on manual reporting, often fails to detect waste accumulation in a timely manner, thereby threatening human health and ecosystem sustainability. As an innovative solution, this study successfully designed and implemented an intelligent waste detection system based on image processing integrated into autonomous unmanned aerial vehicles (drones). Through high-resolution aerial imagery scanning processed with deep learning algorithms, this system is capable of transmitting real-time location coordinates of waste accumulation to relevant authorities using GPS modules and GSM networks. Testing in real urban environments under various conditions demonstrated outstanding performance, with the system achieving a detection accuracy of 89.4%, an average data transmission latency of 2.3 seconds, a scanning area coverage of up to 1.2 km² per flight hour, and a GPS coordinate precision of ±3 meters. With these results, this technological integration has proven to be an effective, efficient, and scalable solution for proactive waste monitoring, which can be further developed in the future alongside robotics-based transportation systems to create a fully automated, responsive, and environmentally friendly urban waste management ecosystem.