Can We Reliably Improve the Robustness to Image Acquisition of Remote Sensing of PV Systems?

TL;DR

This paper investigates the reliability of remote sensing techniques for monitoring rooftop PV systems, introducing a wavelet-based method to enhance robustness against variations in image acquisition conditions.

Contribution

It proposes the WCAM method to analyze and improve the robustness of PV detection models to acquisition shifts, advancing remote sensing reliability.

Findings

WCAM effectively decomposes model predictions to identify scale sensitivities.

The method provides insights to develop more robust PV detection models.

Enhanced robustness increases trust in remote sensing for energy system monitoring.

Abstract

Photovoltaic (PV) energy is crucial for the decarbonization of energy systems. Due to the lack of centralized data, remote sensing of rooftop PV installations is the best option to monitor the evolution of the rooftop PV installed fleet at a regional scale. However, current techniques lack reliability and are notably sensitive to shifts in the acquisition conditions. To overcome this, we leverage the wavelet scale attribution method (WCAM), which decomposes a model's prediction in the space-scale domain. The WCAM enables us to assess on which scales the representation of a PV model rests and provides insights to derive methods that improve the robustness to acquisition conditions, thus increasing trust in deep learning systems to encourage their use for the safe integration of clean energy in electric systems.