A Two-Component AI Framework for Early Warning and Image-Based Diagnosis of Tuta Absoluta in Tomato Crops
Supervisor Name
Dr. Shadia Baroud
Supervisor Email
sy.baroud@alaqsa.edu.ps
University
Al-Aqsa University
Research field
Artificial Intelligence
Bio
Dr. Shadia Yahya Baroud is an Assistant Professor at Al-Aqsa University and holds a PhD in Computing (Artificial Intelligence) from the University Malaysia of Computer Science and Engineering. She is an IEEE Senior Member and a two-time ITU AI for Good Scholar (2024 & 2025). Her research focuses on multi-agent systems, federated learning, generative AI, LLMs and semantic communications for 6G, with applications spanning smart manufacturing, healthcare, intelligent transport and natural language processing. Her recent work includes EcoTransTwin, a federated digital twin framework for sustainable transport decarbonization developed in collaboration with the University of Glasgow and AI-powered multi-agent systems for fraud detection and clinical decision support. She has presented her research at numerous international conferences, including IEEE WCNC, IEEE ICMCTC and the ITU AI for Good Summit in Geneva. Beyond academia, she is committed to expanding access to AI tools, research opportunities and digital infrastructure across underserved and resource-constrained communities and to building AI systems that remain robust and effective under real-world AI architectures.
Description
Tuta absoluta is a destructive tomato pest whose population can increase rapidly before severe plant damage becomes visible. This study develops a complementary artificial-intelligence framework for short-term pest forecasting and image-based diagnosis. The forecasting component used 1,736 surveillance observations from 229 pheromone traps in Malawi, enriched with seven-day historical temperature, humidity, rainfall and wind-speed variables derived from NASA POWER. A chronological split trained the models on 1,498 earlier observations and evaluated them on 138 later observations. Random Forest achieved the best root mean squared error of 324.62 and an R-squared value of 0.238, reducing RMSE by 11.16% relative to a persistence baseline. A MobileNetV2 transfer-learning classifier was separately trained on 6,015 balanced tomato images to distinguish healthy plants, T. absoluta damage, and other tomato diseases. On 903 unseen images, it achieved 79.29% accuracy and a macro F1-score of 79.06%; recall for T. absoluta was 79.07%. Results show that historical trap counts were substantially more informative for next-observation forecasting than the available district-level weather variables, while image classification provided a useful complementary diagnostic signal. The framework supports early warning but requires local, multi-season validation before operational deployment.