<![CDATA[Manual Jar Testing for coagulant dosage determination in water treatment is labor-intensive, time-consuming, and susceptible to operator bias, limiting the ability of Indonesian Regional Water Utilities (PDAM) to respond in real-time to dynamic raw water quality changes. The research contribution of this study is: (1) an ESP32-based automated Jar Testing platform with closed-loop DC motor control and a multi-parameter sensor array (turbidity, TDS, pH) for objective, repeatable coagulation-flocculation evaluation; and (2) a real-time 3D cloud visualization framework using MQTT and Three.js that provides remote monitoring with sub-1.5-second latency. The system integrates a SEN0189 turbidity sensor, a TDS conductivity sensor, and a pH-4502C sensor, each calibrated against laboratory-grade reference instruments using polynomial calibration equations derived from experimental data. Encoder-based closed-loop feedback regulates DC motor speed across a 0–100 RPM range, while all sensor telemetry is transmitted via the MQTT publish-subscribe protocol to a cloud database and rendered by a Three.js-based 3D digital visualization interface. Sensor validation yielded Mean Absolute Errors (MAE) of 0.15 NTU for turbidity, 5.33 ppm for TDS, and 0.04 pH units, all within the respective sensor tolerance bounds. DC motor control achieved MAE of 0.05–0.30 RPM across the 10–100 RPM setpoint range. Six discrete alum dosage trials on raw water with initial turbidity of 6.6 NTU identified 70 mg/L as the optimal concentration, achieving 90.15% turbidity removal with a residual turbidity of 0.65 NTU, below the 1 NTU threshold for potable water quality. The 3D visualization layer-maintained data synchronization latency below 1.5 seconds under laboratory network conditions. The proposed system substantially reduces operator workload and eliminates visual observation bias compared to conventional manual Jar Testing, offering a scalable and low-cost platform for data-driven coagulant dosing optimization in modern water treatment facilities.]]>
