Sense Aeronautics has launched an automated solar inspection capability powered by SenseAI, an advanced image analysis pipeline designed to interpret infrared (IR) and electroluminescence (EL) data captured by drone-based surveys.
Built on the company’s background in aerial AI applications such as object detection, automatic target recognition, and foreign object debris monitoring, the new system applies this expertise to photovoltaic asset management.
AI Aligned with Operational Realities
Rather than positioning artificial intelligence as an isolated element, Sense Aeronautics has structured SenseAI around the realities of aerial data acquisition. The system is designed to work reliably with heterogeneous imagery, reflecting how inspection data is collected and consumed in practice. Outputs are optimized to support maintenance and operations teams by translating complex imaging inputs into consistent, interpretable results.
Infrared Inspection Models
Infrared imaging remains the most widely deployed aerial inspection method for solar panels, used to detect variations in thermal behavior that indicate underlying electrical or mechanical issues. SenseAI’s IR model performs classification at panel level, recognizing that many defects, such as string failures, junction box overheating, and shading effects, that can manifest as distributed thermal patterns rather than discrete objects.
The model architecture is based on YOLO11n-cls, selected for its computational efficiency and suitability for diffuse, panel-wide temperature gradients. Testing demonstrates stable performance across diverse inspection conditions, with high accuracy for high-contrast thermal anomalies including short circuits and substring failures.
Electroluminescence Analysis
SenseAI’s EL model targets micro-level structural defects visible only under electrical stimulation, including microcracks, finger interruptions, dislocations, and black core defects. Built on a YOLOv5s detection backbone, the model identifies and localizes each defect type, enabling visual correlation with ground truth and long-term degradation assessment.
High precision is achieved for well-defined geometries such as dislocations and short circuits, while conservative behavior for subtle, low-contrast defects mirrors how expert inspectors prioritize diagnostic certainty over exhaustive recall.
Integration and Deployment Options
SenseAI is available through a web-based application for turnkey inspections or via an API for integration with existing digital platforms. The API enables automatic image ingestion, analysis, and result retrieval, supporting use within SCADA systems, digital twins, or maintenance management tools.
Deployment flexibility allows the system to operate in cloud environments, secure data centers, or on isolated on-premise infrastructure depending on customer requirements for data governance and connectivity.
Supporting Scalable Solar Operations
By combining operational AI design with multi-modal imaging analysis, Sense Aeronautics has developed a robust foundation for automated photovoltaic inspection. The company continues to refine SenseAI with new datasets and validation feedback to enhance reliability and adaptability across inspection conditions.
Sense Aeronautics invites drone operators, inspection service providers, and platform developers to collaborate in testing and integration of the SenseAI-powered solar inspection system.
The post Automated Solar Panel Drone Inspection Powered by SenseAI appeared first on Unmanned Systems Technology.
