This study proposes a deep learning approach based on the YOLOv8 algorithm for automated tomato disease detection. The model was trained on a custom dataset, which was augmented from an existing Roboflow dataset, achieving an overall accuracy of 66.67%. However, class-specific performance varies, highlighting challenges in differentiating certain diseases. The study suggests further research focusing on data balancing, exploring alternative architectures, and adopting disease-specific metrics. The work lays the foundation for a robust disease detection system to improve crop yields, quality, and sustainable agriculture in Saudi Arabia. Tomatoes are essential in vegetable crop production, with significant yield variations due to factors such as diseases. In Saudi Arabia, tomato cultivation spans 1260 hectares of greenhouse area, exhibiting varied productivity. Tomato harvesting faces challenges with damage rates as high as 10%. Advances in automation technology offer promising solutions, focusing on image recognition, precise positioning, and efficient picking mechanisms. Traditional methods relied on complex machine learning algorithms, but the advent of Convolutional Neural Networks (CNNs) has significantly improved efficiency and precision. Object detection algorithms, crucial for automating tomato disease detection, include two-stage algorithms such as R-CNN and one-stage algorithms such as SSD and YOLO, with various iterations of YOLO improving object detection capabilities. The study used a dataset called "balanceddata Dataset" in Roboflow, published in October 2023, and contained three types of tomato diseases. The dataset was augmented to improve the model performance. YOLOv8 was used as the foundational architecture for the proposed model, as it offers enhanced efficiency and flexibility and addresses three key computer vision tasks: classification, detection, and segmentation. The YOLOv8 design consists of 53 convolutional layers and is implemented as a CNN. It integrates cross-stage partial connections, enabling improved information transmission between layers for enhanced performance. YOLOv8 stands out for its ability to efficiently detect objects at multiple scales because it employs a feature pyramid network. The model was evaluated using precision, recall, and mean Average Precision (mAP) metrics. The confusion matrix showed the model's performance in different types of tomato diseases. The model achieved 66.67% accuracy across all diseases, but there were differences in performance for each disease. The study suggests future work should focus on addressing data imbalance by adding healthy class, data collection or augmentation, exploring alternative architectures or incorporating domain knowledge, and evaluating disease-specific metrics for a more practical perspective. These improvements can pave the way for a robust and reliable automated tomato disease detection system, enhancing crop monitoring and contributing to sustainable agriculture.This study proposes a deep learning approach based on the YOLOv8 algorithm for automated tomato disease detection. The model was trained on a custom dataset, which was augmented from an existing Roboflow dataset, achieving an overall accuracy of 66.67%. However, class-specific performance varies, highlighting challenges in differentiating certain diseases. The study suggests further research focusing on data balancing, exploring alternative architectures, and adopting disease-specific metrics. The work lays the foundation for a robust disease detection system to improve crop yields, quality, and sustainable agriculture in Saudi Arabia. Tomatoes are essential in vegetable crop production, with significant yield variations due to factors such as diseases. In Saudi Arabia, tomato cultivation spans 1260 hectares of greenhouse area, exhibiting varied productivity. Tomato harvesting faces challenges with damage rates as high as 10%. Advances in automation technology offer promising solutions, focusing on image recognition, precise positioning, and efficient picking mechanisms. Traditional methods relied on complex machine learning algorithms, but the advent of Convolutional Neural Networks (CNNs) has significantly improved efficiency and precision. Object detection algorithms, crucial for automating tomato disease detection, include two-stage algorithms such as R-CNN and one-stage algorithms such as SSD and YOLO, with various iterations of YOLO improving object detection capabilities. The study used a dataset called "balanceddata Dataset" in Roboflow, published in October 2023, and contained three types of tomato diseases. The dataset was augmented to improve the model performance. YOLOv8 was used as the foundational architecture for the proposed model, as it offers enhanced efficiency and flexibility and addresses three key computer vision tasks: classification, detection, and segmentation. The YOLOv8 design consists of 53 convolutional layers and is implemented as a CNN. It integrates cross-stage partial connections, enabling improved information transmission between layers for enhanced performance. YOLOv8 stands out for its ability to efficiently detect objects at multiple scales because it employs a feature pyramid network. The model was evaluated using precision, recall, and mean Average Precision (mAP) metrics. The confusion matrix showed the model's performance in different types of tomato diseases. The model achieved 66.67% accuracy across all diseases, but there were differences in performance for each disease. The study suggests future work should focus on addressing data imbalance by adding healthy class, data collection or augmentation, exploring alternative architectures or incorporating domain knowledge, and evaluating disease-specific metrics for a more practical perspective. These improvements can pave the way for a robust and reliable automated tomato disease detection system, enhancing crop monitoring and contributing to sustainable agriculture.