Page 858 - AI for Good Innovate for Impact
P. 858
AI for Good Innovate for Impact
Use Case 5: Edge AI and IoT-Powered System for Early Tomato
Disease Detection and Smart Control
Country: Tanzania
Organization: Mbeya University of Science and Technology (MUST)
Contact Person(s): Yona H. Mgongolwa, yonamgongolwa@ gmail .com, Mr. Zacharia Mzurikwao,
zacerbonnie@ gmail .com
1 Use Case Summary Table
Category Smart Agriculture
The Problem Addressed Delayed detection of tomato plant diseases leads to significant yield
to be addressed reduction and increased economic losses for farmers.
Key aspects of the solu- Deployment of Edge-Artificial Intelligence (AI) for real-time disease
tion detection, integrated with Internet of Things (IoT) - based smart moni-
toring to enable timely and informed crop management decisions.
Technology keywords Edge-AI, IoT, Tomato Leaf Diseases, YOLOv9, Smart Agriculture
Data Availability Open-source image and sensor datasets sourced from platforms such
as Kaggle and Plant Village.
Metadata (type of data) Visual data (images of tomato leaves), Sensor data (temperature,
humidity) collected in real farming environments.
Model Training and YOLOv9 deep learning model trained for accurate tomato leaf disease
fine-tuning classification (achieving 98.6% accuracy); optimized for low-power
edge and mobile deployment.
Testbeds or pilot deploy- Pilot tested on the demonstration farm at Mbeya University of Science
ments and Technology (MUST).
2 Use Case Description
2�1 Description
Agriculture remains a cornerstone of Tanzania’s economy, particularly in the Southern Highlands
region, which includes Mbeya, Njombe, Rukwa, Ruvuma, and Songwe. Among the major crops
cultivated, tomatoes play a critical role in both economic stability and food security. However,
tomato production in these areas faces a persistent threat from diseases such as early blight,
late blight, and septorial leaf spot, which cause substantial yield losses and directly affect the
livelihoods of smallholder farmers.
822
