• Dr.N. S. Ambatkar
• Apeksha Gawai
• Mansi Badwaik
• Navina Likhar
• Bhushan Gautam

Off-grid photovoltaic (PV) installations frequently suffer from premature electrochemical storage degradation resulting from uncoordinated consumer load demand, diurnal insolation intermittency, and recurring deep discharge beneath critical thermodynamic threshold voltages. In sealed valve-regulated lead-acid (VRLA) batteries, repeated discharge below 10.8 V induces irreversible plate sulfation and active paste shedding, curtailing operational service lifespan by up to 70%. To resolve these operational limitations, this paper presents HelioSync, an autonomous cyber-physical edge-IoT platform integrating precision analog instrumentation, internal-resistance compensated State-of-Charge (SOC) estimation, machine learning-driven solar irradiance forecasting, and a deterministic 4-mode energy arbitration state machine. An Espressif ESP32 dual-core 32-bit microcontroller running FreeRTOS acquires multi-channel electrical parameters via calibrated resistive dividers and Allegro ACS712 Hall-effect sensors executing 64-sample circular FIR digital filtering. Multi-variable JSON telemetry is streamed securely over TLS-encrypted MQTT at 2 Hz to an enterprise cloud broker. Concurrently, an edge-to-cloud machine learning pipeline utilizing Random Forest (RF) and Gradient Boosting Regressors forecasts upcoming 30–60 minute photovoltaic yields based on temporal solar angles, lagged harvesting signatures, and ambient environmental metrics (achieving R² = 0.942, MAE = 1.18 W). When battery depletion boundaries are approached, the arbitration finite state machine (FSM) executes proactive load shedding, bounding Depth-of-Discharge (DoD) to 70% while employing a 180-second anti-chatter dwell timer to protect electromechanical relay contacts. Empirical 14-day laboratory bench and outdoor insolation benchmarking confirm that HelioSync extends secondary lead-acid battery cycle life by approximately 2.81× (from ~400 cycles to over 1,120 cycles), while providing comprehensive remote situational awareness at an edge hardware bill of materials (BOM) cost under ₹ 3,780 ($45 USD).

Solar Energy Harvesting, Battery Life Optimization, Artificial Intelligence, Internet of Things (IoT), ESP32, Maximum Power Point Tracking (MPPT), Battery State of Charge (SOC) Estimation, Energy Management, Machine Learning, Predictive Energy Management.

"HELIOSYNC: AN AUTONOMOUS CYBER-PHYSICAL IOT FRAMEWORK FOR REAL-TIME PHOTOVOLTAIC ENERGY HARVESTING, STATE-OF-CHARGE ESTIMATION, AND BATTERY LONGEVITY OPTIMIZATION", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.13, Issue 10, page no.c374-c384, October-2026, Available :http://www.jetir.org/papers/JETIR2610243.pdf

"HELIOSYNC: AN AUTONOMOUS CYBER-PHYSICAL IOT FRAMEWORK FOR REAL-TIME PHOTOVOLTAIC ENERGY HARVESTING, STATE-OF-CHARGE ESTIMATION, AND BATTERY LONGEVITY OPTIMIZATION", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.13, Issue 10, page no. ppc374-c384, October-2026, Available at : http://www.jetir.org/papers/JETIR2610243.pdf