The spread of the moth (Cactoblastis cactorum) in the southern United States, main pest that attacks the Prickly Pear (Opuntia ficus-indica), has raised great concern for the U.S. Department of Agriculture (USDA). The Agricultural Research Service (ARS) and the Animal and Plant Health Inspection Service (APHIS), together with the U.S. Department of the Interior, the Nature Conservancy, and the Secretariat of Agriculture and Rural Development of Mexico are developing strategies to contain the spread of the cactus moth in the region. The use of Artificial Intelligence (AI) tools based on Computer Vision has proven to be a very effective way to detect plant diseases and classify them. This paper proposes an integrated systems composed of a Transfer Learning AI model based on two models YOLOv8n (with window-detection and with segment-detection) and autonomous vehicles (drones and ground robots) to prevent and mitigate the pest attack. The models were chosen based on their efficiency and portability, since they are going to be applied in Android and Linux-based edge devices. The AI models will be evaluated based on the following metrics: precision, recall, accuracy and F1 score; and the overall system will be tested in the field by the end of the year in the US and Brazil. The objective of this work is to demonstrate the use of AI tools for plant health detection on edge devices integrated with autonomous vehicles.