T-Pest: Smart Agriculture with AI, IoT, and GIS for Rice Pest Control

T-Pest: AI-Powered solution for early pest detection

TMA Solutions has developed the T-Pest system, integrating AI, IoT, and GIS technologies to identify 7 types of rice diseases and monitor 8 harmful insects in Bình Định Province. This innovation stems from the research project titled “Application of Artificial Intelligence in Identifying and Monitoring Pests and Diseases in Rice Cultivation.”

Combining the strengths of artificial intelligence, IoT, and cutting-edge information technology, T-Pest aims to optimize pest management, improve crop yield, and modernize agricultural practices in Vietnam.

A Comprehensive solution for smart agriculture

T-Pest is equipped with an automatic data collection device featuring a high-resolution camera, UV light, and environmental sensors. Images collected are analyzed using the enhanced YOLOv5-Ghost AI model, ensuring fast and accurate identification of pests and diseases.

This data is seamlessly integrated into a web-based platform, offering real-time monitoring and visually mapping pest infestations. The system empowers users with actionable insights to combat agricultural challenges.

Key Features of T-Pest:

• Accurate detection: T-Pest can currently identify 7 rice diseases, including bacterial blight, rice blast, brown spot, physiological yellowing, sheath blight, sterile grain, and bacterial leaf streak, as well as 8 harmful insects, such as white-backed planthopper, brown planthopper, stem borer, small leaf roller, rice gall midge, green leafhopper, black bug, and green stink bug.
• Real-time Monitoring: Operates 24/7 in real-time, automatically collecting data and issuing timely pest alerts.
• User-friendly platform: Farmers can access a web-based dashboard to review pest and disease information, receive alerts, and implement suggested solutions promptly.

AI-powered precision

According to Mr. Trần Hoàn Anh Nguyên, Director of TMA’s Data Science Center, the development of T-Pest began with building a comprehensive database of common rice pests and diseases in local fields. Data was gathered using specialized equipment and smartphones, followed by detailed labeling and validation to ensure the accuracy of AI training datasets.

Mr. Nguyễn Quốc Dương, a core researcher, highlighted that optimizing the YOLOv5 AI model significantly enhanced T-Pest’s performance. By reducing the model's size and resource requirements while maintaining high accuracy, the system is well-suited for practical deployment in real-world farming environments.

During field trials conducted in Phù Cát and Tuy Phước districts, T-Pest analyzed over 1,000 images of pests and diseases with an accuracy rate exceeding 80%. The accompanying mobile application further simplifies the process, enabling farmers to quickly assess field conditions, receive alerts, and take timely preventive measures.

A farmer from Phước Sơn Commune, Mr. Nguyễn Văn Chín, shared: "The T-Pest system makes it easier for us to monitor our fields and detect pests without frequent visits. It’s a reliable tool for managing crops and securing better harvests."

Paving the way for agricultural modernization

T-Pest has not only earned recognition from farmers but also from agricultural management authorities. It marks a significant milestone for smart farming initiatives in Bình Định Province and opens new opportunities for high-tech applications in agriculture across Vietnam.

As shared by Mr. Nguyên, TMA Solutions plans to expand T-Pest’s reach beyond Bình Định after the pilot phase. The goal is to enhance pest control measures, boost yields, and safeguard crops, contributing to Vietnam’s vision of sustainable agricultural development.

By integrating AI and IoT technologies, T-Pest demonstrates the transformative power of artificial intelligence in addressing modern agricultural challenges, offering a scalable and effective solution for pest management.