SKRIPSI DIGITAL
Rancang Bangun Sistem Deteksi Kebakaran Berbasis TinyML pada Sensor IR Photoelectric dengan Klasifikasi Tingkat Asap dan Kendali Pompa Otomatis Menggunakan ESP32-CAM = Design and Development of a TinyML-Based Fire Detection System Using a Photoelectric IR Sensor with Smoke Level Classification and Automatic Pump Control Using the ESP32-CAM
This study designs and implements an intelligent early fire detection system based on Edge Intelligence using the ESP32-CAM module and an IR photoelectric sensor. The main issues addressed are the reliance of conventional detection systems on cloud connectivity and the high rate of false alarms. This system employs TinyML technology to classify smoke density levels offline and in real-time directly on hardware with limited memory. The TinyML model was trained using the Edge Impulse platform with raw image datasets from the ESP32-CAM categorized into three labels: Normal, Light Smoke, and Thick Smoke. The system integrates an IR Photoelectric sensor as a dual validation mechanism to improve detection accuracy under dense smoke conditions. When the classification indicates a “Thick Smoke” status and the IR sensor is triggered, the ESP32-CAM automatically activates a buzzer and the fire suppression pump system via a relay module. In addition to automatic control, this system is equipped with an IoT-based remote monitoring feature using the MQTT protocol, integrated with an Android (Flutter) application to display sensor status and classification results in real-time. The results of this research are expected to provide a responsive, energy-efficient, and autonomous fire protection system solution that does not rely entirely on complex internet network infrastructure.
Tidak tersedia versi lain
Karya Umum 
Filsafat 
Agama 
Ilmu-ilmu Sosial 
Bahasa 
Ilmu-ilmu Murni 
Ilmu-ilmu Terapan 
Kesenian, Hiburan, dan Olahraga 
Kesusastraan 
Geografi dan Sejarah