Pengembangan Sistem Smart BMI Berbasis ESP32 untuk Monitoring Pencapaian BMI Ideal

Permasalahan berat badan tidak ideal menjadi salah satu isu kesehatan yang terus meningkat, sedangkan pemantauan Body Mass Index (BMI) masih banyak dilakukan secara manual dan belum disertai rekomendasi aktivitas harian yang dipersonalisasi. Penelitian ini bertujuan mengembangkan Sistem Smart BMI berbasis ESP32 untuk melakukan monitoring pencapaian BMI ideal melalui pengukuran berat badan dan tinggi badan, perhitungan dan klasifikasi BMI, serta pemberian rekomendasi aktivitas harian. Penelitian menggunakan metode Design Science Research (DSR) dengan sistem yang dikembangkan menggunakan ESP32, sensor loadcell, sensor ultrasonik HC-SR04, metode Threshold-Based Classification untuk klasifikasi BMI berdasarkan standar Kementerian Kesehatan Republik Indonesia, serta Gemini API untuk menghasilkan rekomendasi aktivitas harian sesuai kondisi dan target pengguna. Hasil penelitian menunjukkan bahwa sistem berhasil dikembangkan dan seluruh fungsi utama berjalan sesuai rancangan. Sensor loadcell memperoleh akurasi 97,7%, sensor HC-SR04 memperoleh akurasi 99,2%, sedangkan perhitungan dan klasifikasi BMI memiliki tingkat kesesuaian 100% terhadap perhitungan manual. Dengan demikian, sistem yang dikembangkan mampu mendukung monitoring perkembangan BMI secara otomatis serta memberikan rekomendasi aktivitas harian yang dipersonalisasi sesuai kondisi dan target pengguna.

================================================================================

The problem of unhealthy body weight is a growing health issue, yet Body Mass Index (BMI) monitoring is still largely done manually and does not include personalized daily activity recommendations. This study aims to develop an ESP32-based Smart BMI System to monitor progress toward an ideal BMI through measurements of weight and height, BMI calculation and classification, and the provision of daily activity recommendations. The study employed the Design Science Research (DSR) method. The system was developed using an ESP32, a load cell sensor, an HC-SR04 ultrasonic sensor, a threshold-based classification method for BMI classification according to the standards of the Ministry of Health of the Republic of Indonesia, and the Gemini API to generate daily activity recommendations tailored to the user’s condition and goals. The results showed that the system was successfully developed and all main functions operate as designed. The load cell sensor achieved 97.7% accuracy, the HC-SR04 sensor achieved 99.2% accuracy, while BMI calculations and classification showed 100% agreement with manual calculations. Thus, the developed system was capable of supporting automated monitoring of BMI progression and provides personalized recommendations for daily activities based on the user’s condition and goals.