Keeping the surface of photovoltaic (PV) panels clean has been a challenge. The fundamental understanding in interactions between a brush, dust particles, and the surface of panels is still lacking. This research investigates the correlation between surface morphological parameters and cleaning efficiency. Experimentally, the cleaning efficiency using brush tips sweeping against five aluminum oxide surfaces was evaluated. We analyzed the cleaning conditions such as the surface roughness and the shape of the surface profiles. Results showed that there are two key factors affecting the removal of dust particles resulting effective cleaning. The first is the roughness of the surface and the second the profile. The dust particle removal involves consideration of the brush and particle size, as well as the panel surface morphology and texture. A physical model illustrating the cleaning mechanism is proposed. The impact of the research contributes to the surface design leading to the ultimate cleaning efficiency of PV panels.