AI-Powered Smart Farming & Monitoring
The Project
This Project highlights an innovative precision agriculture system integrating an adaptive AI model with soil sensors and a real-time web platform. At the heart of this system is the Synesthetic Fluid Neural Network (SFNN)—a specially designed PyTorch-based deep neural network leveraging modality-specific projection, non-Newtonian dynamic layers, rheological activation, and cross-modal synesthetic coupling to robustly classify crop types based on multi-environmental inputs. The model was trained on a synthetic dataset with careful attention to imitating realistic agro-environmental conditions, including soil characteristics, atmospheric conditions, and environmental conditions. Input data is collected using a portable, low-cost sensing device with a Raspberry Pi 4B, 2 soil sensors, and a GSM module for SMS telemetric data transfer, integrated in a 3D-printed casing. Soil probing is automated, and sensor readings are sent to a remote database. The results are presented in an interactive React + Three.js web dashboard, allowing real-time examination and AI-driven crop prediction within a geospatial 3D space. The framework provides a closely integrated pipeline from physical perception to intelligent inference and user interaction.
