Pengembangan Smart Trashbot Berbasis Internet Of Things Dengan Fitur Gesture Recognition Dan Autonomous Movement
Abstract
This study aims to develop a prototype of the Smart TrashBot, an Internet of Things (IoT)-based mobile smart trash bin, to improve efficiency and hygiene in waste management. Conventional, static, manual trash bins are considered unhygienic and suboptimal. Therefore, this system is designed by integrating gesture recognition features using the ESP32-CAM module to automatically open the lid without physical contact, along with autonomous movement utilizing the front ultrasonic sensor for independent navigation to avoid obstacles. The system is controlled by an ESP32 Dev Board microcontroller and communicates wirelessly via the MQTT protocol. Functional testing using the Black Box method shows that the camera's 3-frame sequential validation algorithm accurately recognizes gestures and successfully reduces false positives. Furthermore, the robot is able to maneuver to avoid obstacles at a distance of less than 25 cm, and effectively sends real-time telemetry notifications to the dashboard system when the remaining trash capacity is below 7 cm. Although slight deviations in the wheel trajectory were observed due to the open-loop control system, overall, the separation of computation between Edge AI and IoT actuator control on the Smart TrashBot successfully creates an interactive, automated, and hygienic waste management solution.

