The use of Unmanned Aerial Systems (UAS) in invasive species management has gained traction in recent years, yet real-time, UAS-based identification of noxious weeds remains a challenge. The development of a Convolutional Neural Network (CNN) to identify Phragmites australis leads towards real-time detection and management of invasive weeds. Using high resolution UAS imagery provides sufficient data to train an accurate CNN model to identify instances of Phragmites australis. The trained CNN model can then be refined and deployed on an edge device for operations on a UAS-based sprayer.