A university team built an AgTech risk monitoring system for a Hardware Architecture course, focusing on real-time environmental data collection. The system uses Arduino with sensors to track temperature, humidity, and luminosity, feeding data into an asynchronous pipeline. This setup enables continuous monitoring and risk analysis for potential pest outbreaks in agricultural settings.

The architecture relies on RabbitMQ for messaging, decoupling hardware from backend services built with Python and Flask. This choice handles network issues common in rural areas. Data is cached in Redis and stored in SQLite, while a Next.js dashboard provides real-time visualization. This modular design separates hardware acquisition from processing and frontend display.

Challenges included ensuring reliable sensor ingestion and balancing real-time needs with analytical flexibility. The team addressed this by using an event-driven pipeline that preserves raw data, allowing for future machine learning integration without altering ingestion. The system proved successful, validating the full data flow from sensors to dashboard in a live demonstration.

Future work involves developing predictive models for pest outbreaks and refining the data pipeline for better quality. Enhancements may include new sensors and improved alert mechanisms. This project serves as a practical foundation for scalable, distributed AgTech solutions, bridging hardware architecture with data engineering and web development concepts.