<![CDATA[The problem of diabetes mellitus (DM) has reached an alarming level, marked by the number of sufferers that continues to increase every year. Based on data from the International Diabetes Federation (IDF), there are around 589 million sufferers worldwide. In Indonesia in 2024, of a total of 185 million adults, the prevalence of diabetes reached 11.3% or around 20.4 million cases. Non-invasive blood sugar monitoring has long been a dream in diabetes management due to the convenience factor, especially for patients with high glucose. This study focuses on monitoring glucose and cholesterol levels non-invasively, reducing discomfort, pain in patients, and a preventive tool in dealing with potential emergencies. The method used in this research is the MAX30102 Sensor which is useful for implementing a multi-wavelength photoplethysmograph on a bracelet that will detect glucose levels through the patient's skin. Data will be channeled and processed by ESP32, OLED LCD as a visualization tool for the processed data, and Wi-Fi connectivity for providing data on a spreadsheet with continuous monitoring to health services. Testing was conducted on 10 respondents, comparing the results between a non-invasive method (NonivaBand) and an invasive method (conventional glucometer). The analysis showed an average error of 5.67% and an accuracy of 94.33%, indicating a strong correlation between the PPG signal and reference blood glucose levels. Future research is expected to use sensors with higher accuracy and more test data to improve measurement results.]]>
