The global decline in natural pollinators, driven by factors such as climate change, habitat destruction, and pesticide use, poses a significant threat to agriculture and food security. This study addresses that challenge by developing a robotic pollination system enhanced with machine learning to support the cultivation of crops like strawberries that rely on effective pollination. A dataset comprising images of strawberry flowers under diverse environmental conditions was assembled to train a detection model. Using a pre-trained YOLO network and camera calibration methods, the system identifies flower stigmas and maps their pixel positions to real-world coordinates with high accuracy. A three-degree-of-freedom robotic arm, fitted with a pollination-specific end effector, executes targeted movements based on inverse kinematics to perform the pollination task. The model’s performance, evaluated using accuracy, precision, recall, and F1-score, confirms its robustness and reliability. This research demonstrates a practical solution that integrates AI and robotics to mitigate pollinator loss, offering a scalable, sustainable tool for improving productivity across various farming systems.