<![CDATA[The Vokafit K1 UNESA GYM building frequently experiences power trips whenever four treadmill units are operated simultaneously, while facility managers lack tools to monitor loads or manage energy consumption preventively. This study designs and tests an Internet of Things (IoT)-based power monitoring and analysis system capable of monitoring four treadmill load points simultaneously while presenting real-time operational cost estimates. The system is built with a Master-Slave architecture: an ESP32 microcontroller reads four PZEM-016 sensors via the Modbus RTU (RS485) protocol, converts consumed energy (kWh) into cost estimates based on the applicable Basic Electricity Tariff (TDL), sends data to the ThingSpeak platform via the MQTT protocol, and backs up data to a MicroSD module when the internet connection is lost. Test results show that the PZEM-016 sensors have voltage, current, and active power deviations below 1% compared with a True RMS multimeter; data on the ThingSpeak dashboard and MicroSD are consistent with 0% deviation; and the cost calculation algorithm in the ESP32 firmware has a maximum error of 0.05%. Cumulative testing over 15 working days recorded net energy consumption of 113.89 kWh with a total operational cost of Rp193,559.47. This study also identifies inrush current as the cause of MCB tripping when several treadmills are turned on simultaneously at high power. Overall, the developed system is proven accurate, reliable for long-term monitoring, and suitable as a basis for energy cost transparency and electrical load management in the Vokafit K1 UNESA GYM building.]]>
