GlobsitSmartDrone, autonomous aerial inspection system, allows accurate analysis of the photovoltaic system through thermographic and multispectrographic detection of anomalies and detailed analysis of the collected data.
The GlobsitSmartDrone aerial inspection service allows you to read the serial numbers of each individual panel, with relative positioning and georeferencing, and to monitor the productivity status of the individual panels. Artificial intelligence is the core GSD; it is in fact a machine learning system trained to guarantee reliability and precision for the analysis and classification of faults and anomalies.
The on-board payload includes very high resolution sensors that are able to correct possible false negatives and positives present during an inspection.
The latest generation thermal sensor, with 640×320 pixel output, admits a greater identification of the number of anomalies and a precise classification of these. It is a high resolution sensor for reliable detections even from a certain distance from the panel.
Thermal images, obtained at the end of the data processing, are analyzed with interpolation of the data coming from the visible sensor, an 8 mega pixel camera with parallax correction to avoid perspective distortions.
Finally, the analysis of the multispectrographic data ensures the most precise possible discrimination of the fault categories. For example, one of the spectra used is the NIR sensor, for a full-details output interpretation.
The entire IT process is governed by two neural networks: a Computer Vision Guided data collection system and a second neural network that manages the entire analysis phase, thus providing indications for predictive maintenance, for increasing the performance of the plant and for the optimization of maintenance costs. The Computer Vision uses sensors and 6 cameras on visible on-board and lets the maximum autonomy of overflight of the SAPR without setting plans in waypoints and static routes. GlobsitSmartDrone is therefore able to fly over strings at close heights with real-time analysis regardless of operational scenarios. The second neural network mash-ups the collected data from the analysis of which it is possible to draw information on the productivity status of the panel and possible failures.
The GlobsitSmartDrone AI system has been specifically designed to calculate the degradation of the panels in percentiles and provide the parameters for predictive and traditional maintenance.
In addition to reporting faults, GSD has been designed for detecting freshmen. The survey service also guarantees maximum precision in the selection and positioning of the serial numbers. The Drone or Ground Rover (UAV / ULV) is able to read the data from the front label affixed to the front or back of the photovoltaic module and perform a careful comparative evaluation between barcodes and numerical formats. In this way, an accurate reading is guaranteed even in the presence of spots and possible burns on the labels. Reading anomalies are managed by the second neural network, in the latter case, the images are reread for a second time until an analysis of the data is reached as reliable as possible.
Thanks to AI automatisms, Globsit ensures a margin of error in inspection and analysis of 1%.
All data relating to the operation of the photovoltaic panels, once merged into the servers and analyzed, will be made available to the customer in a cloud solution, which can be customized and downloaded from a computer, tablet and telephone.
The GSD-Console, the cloud software designed to help our customers monitor and improve the performance of their photovoltaic systems, therefore allows the customer to evaluate and schedule maintenance effectively, making controlled replacement and targeted cleaning interventions cost-effective. A georeferenced map in fact indicates the correct position of each individual module analyzed and is able to provide correct information to the customer.
Finally, from the Dashboard oin cloud, it is possible to download a pdf-report with the statistics for each panel, string and fault, in criticality levels of the anomalies and data on temperature and equipment.
GSD Dashboard provides the tools to synchronize maintenance operations, including the information necessary for scheduling cleaning interventions and constant monitoring for plant safety.
The inspections on photovoltaic systems carried out by GlobsitSmartDrone respects the IeC TS62446-3 Standard on external thermographic investigations of photovoltaic modules and systems, under certain operating conditions. Based on this, GlobsitSmartDrone relies on geometric thermographic resolutions with a minimum resolution of 5×5 pixels for each individual solar cell to obtain a high quality image.