Birds are widely known biological indicators for ecosystem health.  In conjunction with the efforts to transition the U.S. to clean energy, there is uncertainty about how birds respond to associated infrastructure, and what those responses may mean to us. To overcome limited observational capability, we developed a bird monitoring camera system using an edge-computing video camera and a suite of machine learning (ML) models to continuously observe bird activities at photovoltaic solar facilities. Each of the four ML models in our AI suite accomplishes moving object detection, bird detection, bird collision detection, and non-collision activity classification, respectively. The latest field validation confirmed accuracies of 94% bird detection, 81% collision detection, and 83% non-collision act