The rapid growth of solar energy installations worldwide calls for innovative solutions to optimize the operations and maintenance (O&M) activities in solar energy farms, with the ultimate goal of enhancing the economic outlook of solar power. Recently, there has been a growing interest in exploring the merit of emerging technologies such as unmanned aerial vehicles (UAVs) and artificial intelligence (AI) in driving smart O&M decisions for solar photovoltaic (PV) systems. Towards this goal, this paper presents a UAV-enabled, AI-powered framework to automate solar energy asset monitoring and fault detection. First, an experimental testbed has been set up at the Energy Lab at Rutgers University – New Brunswick, wherein a UAV is flown over an operational PV system to collect real-time, high-resolution aerial images of the solar panels under various operational and weather conditions. Then, a deep-learning (DL)-based framework is proposed to extract relevant features from the processed UAV images, which are then combined with exogenous weather parameters, in order to make a decision on the health status of the solar panel under inspection. Our extensive experiments on two prevalent fault modes, namely snow accumulation and shading, suggest that our proposed approach can effectively identify the occurrence of such faults in solar panels, with up to 95.6% accuracy, while maintaining a sensible balance between false and missed alarms. Our framework serves as a testament to the merit of combining UAV-enabled data acquisition with emerging AI technologies in order to automate and optimize O&M activities and asset management in solar farms.