“…However, some challenges may arise when developing such systems, deteriorating the effectiveness and the autonomy of UAV's. Such challenges are concerned with the localization of the drone (where the precision of the geolocalization is not sufficient to allow flying at low altitudes), autonomous flight duration and flying paths over the plant to avoid overlap and empty areas between the PV strings, acquired images resolution and quantification of the impact of each fault's pattern on PV performance degradation [ 64 , 72 , 115 , [134] , [135] , [136] , [137] , [138] ]. Towards tackling these challenges, vision-based control laws were suggested to track PV panel rows based on PV modules' edge detection [ 134 , 136 , 139 ], while different techniques were also proposed to extract the plant's boundary via computer vision techniques and compute the UAV path over the plant [ 135 , 138 ].…”