<![CDATA[Abstract— The problem faced in the biogas fermentation process is the challenge of continuously monitoring environmental conditions such as temperature, humidity, methane gas (CH?) concentration, and pressure, which have a major effect on gas production efficiency. This research aims to design a biogas fermentation monitoring system that uses Internet of Things (IoT) technology so that it can automatically classify fermentation conditions with the help of the K-Means Clustering algorithm. The system utilizes ESP32 microcontroller connected with DHT22 and MQ-4 sensors to measure temperature, humidity, and CH? parameters, and sends the data directly to Blynk platform via WiFi connection. The data collection process was carried out every five hours for 15 days after the initial fermentation lasted for three weeks. The resulting data was then analyzed using the K-Means algorithm to classify fermentation conditions into three categories: early, transitional, and active. Evaluation results using the Elbow and Silhouette Score methods indicated that the ideal number of clusters was three (K=3), with most of the data belonging to the active cluster. The 3D representation and scatter diagram confirmed that each cluster had significantly different sensor characteristics. The system successfully facilitated the monitoring of the fermentation process and provided important classification information to support decision-making. This research shows that combining IoT and machine learning can improve the efficiency of biogas fermentation management.]]>
