Follow India Renewable Energy News on WhatsApp for exclusive updates on clean energy news and insights

NIT Rourkela Develops AI-Powered Smart Monitoring System to Boost Solar Plant Performance and Reduce Cleaning Costs

Researchers at the National Institute of Technology (NIT) Rourkela have developed and patented an Artificial Intelligence (AI)-based autonomous monitoring system that can detect faults, assess panel soiling and optimise cleaning schedules for solar power plants. The technology, based on federated learning, is designed to improve photovoltaic (PV) plant efficiency while reducing maintenance costs, water consumption and unnecessary manual interventions.

The patented innovation, titled "Federated Learning Based Autonomous System and Method for Monitoring and Cleaning Solar Plant," continuously monitors the operating condition of solar panels and identifies dust accumulation, performance degradation and potential equipment faults in real time. By analysing operational data through AI algorithms, the system determines when and where cleaning or maintenance is actually required, allowing operators to shift from routine maintenance schedules to a more efficient condition-based maintenance approach.

A distinguishing feature of the technology is the use of federated learning, an advanced machine learning technique that enables multiple devices or sites to collaboratively train AI models without transferring raw operational data to a central server. Instead of sharing sensitive plant data, only the learning parameters are exchanged, improving data privacy and cybersecurity while enabling continuous model improvement across geographically distributed solar plants. This approach is particularly valuable for utility-scale solar developers managing large renewable energy portfolios spread across different locations.

Soiling caused by dust, sand, bird droppings, pollution and other environmental factors remains one of the biggest operational challenges for photovoltaic power plants. Depending on climatic conditions, dust accumulation can reduce solar energy generation by 5-30 percent or even more if panels are left uncleaned for extended periods. Conventional cleaning practices generally rely on fixed schedules, which often lead to unnecessary cleaning, higher labour costs and excessive water consumption, especially in water-scarce regions.

The AI-enabled system developed by NIT Rourkela addresses this challenge by identifying only those panels or sections experiencing significant soiling or abnormal performance. This targeted approach enables operators to optimise cleaning frequency, reduce operational expenditure (OPEX), minimise water usage and improve overall plant availability without compromising electricity generation. The system can also support predictive maintenance by identifying equipment anomalies before they develop into major faults, helping reduce downtime and maintenance expenses.

The innovation comes at a time when India's solar sector is expanding rapidly under the country's clean energy transition programme. As India works towards achieving 500 GW of non-fossil fuel-based installed power capacity by 2030, utility-scale solar parks are becoming larger and more geographically dispersed, increasing the need for intelligent digital asset management solutions. Artificial intelligence, machine learning, robotics, drones and Internet of Things (IoT)-based monitoring systems are increasingly being deployed to improve plant performance, automate inspections and lower lifecycle operating costs.

AI-driven operation and maintenance technologies are emerging as an important component of next-generation renewable energy infrastructure. Combined with robotic cleaning systems, digital twins, predictive analytics and automated fault detection, such innovations can significantly improve energy yields while extending the operational life of solar assets. The integration of privacy-preserving AI techniques such as federated learning further enhances the scalability of these solutions across large renewable energy portfolios.

The patented technology developed by NIT Rourkela demonstrates the growing role of Indian research institutions in advancing indigenous clean energy technologies. By combining artificial intelligence, autonomous monitoring and data-secure machine learning, the innovation has the potential to improve the operational efficiency of solar power plants, reduce resource consumption and support the long-term sustainability of India's rapidly expanding renewable energy sector.