Design and Development of a Solar-Powered IoT-Based Energy Consumption Monitoring System

Abstract

Residential energy conservation is frequently hindered by a lack of real-time visibility into usage, widening the "behavior-action gap" between daily electricity utilization and delayed monthly utility statements. This paper presents the development of a low-cost, open-source, residential solar-powered Internet of Things (IoT) monitoring system. The proposed architecture segments data collection tasks into a Direct Current (DC) renewable generation node and an Alternating Current (AC) appliance-level consumption interface using dual ESP32 microcontrollers. The system collects data from precision inline INA226 shunt sensors on the solar and battery storage paths, alongside a ZMPT101B transformer and an SCT-013 current transformer wrapped in a physical 5-turn wire loop to lift low-power current boundaries above standard resolution floors. Telemetry is streamed asynchronously via Message Queuing Telemetry Transport (MQTT) over a cloud HiveMQ broker. To optimize edge node efficiency, chronological timestamping is completely decoupled from the hardware layer and managed server-side upon database ingestion into InfluxDB. Real-time and historical analytics are visualised using a centralized Grafana web dashboard interface. Experimental results show that the hardware sensing layer maintained high measurement precision, yielding low absolute errors (Verror ≤ 0.38% and Ierror ≤ 32%) and successfully resolving low-power household standby states down to single-digit wattages (e.g., 9W). The event-driven alert system achieved zero-latency multi-tier warnings, verifying that the proposed framework delivers an accessible, high-fidelity sub-metering platform for urban electrification programs.