Solar optimization via learning-driven visual intelligent analytics

The accumulation of airborne particulate matter on photovoltaic modules significantly reduces energy yield underscoring the need for intelligent and cost-effective cleaning strategies. Despite progress in soiling mitigation current approaches remain constrained by rigid cleaning schedules, reactive maintenance and limited integration of predictive modeling with economic optimization. To bridge these gaps, this paper introduces Solar Optimization via Learning-driven Visual Intelligent Analytics (SOLVIA), a globally adaptable end-to-end framework for predictive cleaning and intelligent maintenance of photovoltaic systems. SOLVIA unifies heterogeneous data sources including optical soiling sensors, particulate matter (PM2.5/PM10) detectors, irradiance meters, temperature probes, wind speed sensors and publicly available environmental APIs into a regression-driven predictive engine that forecasts dust accumulation, power degradation and financial losses. Cleaning interventions are dynamically recommended when projected economic losses surpass threshold constraints ensuring that maintenance actions are both timely and cost-effective. A cloud-ready web application delivers interactive dashboards with geospatial insights, performance forecasting and cost-benefit decision support, thereby transforming complex analytics into actionable strategies. The framework was deployed and validated on a 24.12 kW rooftop solar PV installation at the Netaji Subhas University of Technology, Dwarka, Delhi, India, a high-dust region where soiling severely affects system performance. Results demonstrate predictive accuracy with RMSE = 3.8 W, MAE = 2.6 W, R2 = 0.95 enabling optimized cleaning schedules that reduced unnecessary interventions by over 30 % compared to fixed-interval policies. Furthermore, SOLVIA's economic optimization workflow was evaluated across four geographically diverse sites Delhi, Srinagar, Bonn, and Melbourne demonstrating strong adaptability to variable dust loads, climatic regimes, and operational conditions. These outcomes highlight SOLVIA's ability to enhance energy yield, minimize maintenance costs and scale across diverse geographic regions establishing it as a robust and economically sustainable solution for solar asset management worldwide.