Photovoltaic power plants play a crucial role in the global energy transition, but their efficiency depends heavily on continuous monitoring and timely maintenance. A single damaged or dirty panel can reduce energy output across an entire array, making early fault detection essential for large-scale solar farms as well as smaller distributed installations.

Conventional inspection methods are largely manual or semi-automated, requiring technicians to operate drones or use handheld thermal cameras. These approaches are time-consuming, expensive and limited in coverage. They also rely on human interpretation of data, which often delays diagnostics and increases the risk of undetected faults.

Environmental conditions such as high temperatures, dust, or uneven terrain add further complexity. Regular cleaning and inspection are critical but difficult to coordinate efficiently. Without automation, maintenance schedules are reactive rather than predictive, leading to avoidable downtime and long-term performance degradation.

The system introduces a fully autonomous approach to monitoring and maintaining photovoltaic power plants. It combines advanced drone technology with artificial intelligence to automate inspections, diagnostics and cleaning. Equipped with high-resolution visual and thermal sensors, the drones detect dust, cracks or damaged modules in real time, identifying performance issues before they affect energy output.

A dedicated cleaning drone complements the inspection process by performing targeted maintenance only where needed. This smart coordination minimizes water and energy use while keeping panels in optimal condition. The entire fleet operates autonomously, guided by an adaptive mission planner that adjusts flight routes based on site layout, weather conditions and detected faults.

All components are integrated into a secure digital platform that processes and stores collected data for predictive maintenance. AI algorithms analyze performance trends, prioritize interventions and automatically generate maintenance schedules. The result is a sustainable, cost-effective system that improves long-term reliability and strengthens the efficiency and competitiveness of solar energy infrastructure.

Our solution combines robust hardware, advanced autonomy, and flexible sensor integration to turn raw inspection data into meaningful insights that improve safety, efficiency, and decision-making.

• The platforms are designed to operate both in GNSS and GNSS-denied environments, featuring advanced localization, obstacle avoidance, and fully autonomous flight trajectories.

• The mission-planning interface allows operators to define inspection points, automate scan patterns, and monitor telemetry in real time, including battery status, sensor data, and communication link quality.

• A modular design supports various sensor configurations (thermal cameras, LiDAR, optical or ultrasonic sensors), customizable for specific inspection needs such as structural integrity, pipelines, or machinery.

• Multi-robot coordination enables cooperative missions, for example, simultaneous scanning of interior and exterior areas, synchronized data acquisition, and faster coverage of large sites.

• The user interface provides straightforward mission setup, result visualization, and data interpretation tools, transforming every inspection into a source of actionable operational intelligence.

See the Project Outcomes

Algorithms for automatic visual defect detection

HW functional sample of unmanned helicopter with integrated flight control unit.

SW for autonomous inspection via unmanned helicopters.

User interface SW for specifying inspection missions, interaction with an autonomous helicopter and visualization of flight and inspection data.

Prototype of an unmanned helicopter optimized for performing autonomous inspection missions in complex environments

Prototype of an unmanned helicopter optimized for performing autonomous inspection missions in complex environments

Functional sample of drone flight control unit.