top of page

Real-time defect detection in PV panels on unmanned aerial vehicles (UAV) devices

Use case

The “Real-time Fault Detection in PV Panels on Unmanned Aerial Vehicles (UAVs)” is one of the REBECCA project’s Use Case that is being developed by Intecs Solutions. The goal of this Use Case is to develop an automated Deep Learning-based model for defect detection in PV panels based on aerial images captured by IR camera mounted on board of an Unmanned Aerial Vehicle (UAV) and with the processing executed directly on-board.

The results of this processing will be displayed as a predicted on-ground defect positions in a 2D map of a PV plants directly visible on the remote controller of the UAV pilot. The current state of the art for AI-based solutions in this area typically involves initial data collection by O&M operators in the field, followed by offline processing in the O&M control room via a workstation.

This results in a delayed intervention by the O&M operators, which leads to higher maintenance costs. Example of O&M OperationsIntroducing on-board processing bypasses the need for data transfer and enables instant analysis, reducing O&M costs while increasing maintenance effectiveness.

Our goal in this case study is to streamline the model pipeline to reduce inference time, develop leaner CNNs, and employ structured/unstructured pruning techniques to reduce the required volume and floating-point operations (FLOPs) while ensuring a good trade-off with accuracy metrics. We're also exploring the integration of GNSS receivers to increase the resolution of defect localization to single cell granularity.

In this phase we are currently working on our algorithm and in defining collaboration with the REBECCA’s subcomponents providers.

www.intecs.it

ACKNOWLEDGMENT: REBECCA project is supported by the Chips Joint Undertaking and its members, including the top-up funding by National Authorities under grant agreement n° 101097224. Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the granting authority. Neither the European Union nor the granting authority can be held responsible for them.
EN_Co-fundedbytheEU_RGB_WHITE.png
chips_white.png
Logo_white.png
  • X
  • LinkedIn
  • Youtube
bottom of page