Agritech Harmony: Real-time ESP32 Automation and Cloud Computing for Socially-Informed Crop Suggestions
Vishal Bawadkar, Aman Verma, Aditya Pawar, Utkarsh Phatale, Prashant Sadaphule · 2024
Accurately predicting long-term crop yield trends remains a crucial challenge in optimizing agricultural practices and ensuring food security. This paper proposes a novel frame-work that merges real-time data acquired from Internet of Things (IoT) sensors with a Bi-directional LSTM model with an attention mechanism to achieve this objective. This architecture extracts insights from historical and future environmental conditions and prioritizes critical factors influencing yield through attention weighting. The framework incorporates an ESP32-managed sensor network measuring humidity, temperature, moisture, rainfall, and NPK levels, providing a continuous stream of real-time field data. Evaluation of real-world datasets demonstrates the framework's superiority over traditional methods, yielding significantly higher prediction accuracy and exposing key drivers of yield fluctuations through attention analysis. An accessible API facilitates seamless integration with farm management systems, empowering farmers with data-driven decision-making capabilities. This paper highlights the transformative potential of integrating deep learning, real-time IoT data, and APIs for revolutionizing agriculture, paving the way for a more sustainable and resilient agricultural futur