RESEARCH PROJECT 01
SMART FASAL
AI-driven agricultural research integrating Internet of Things, environmental sensing, data analysis and machine learning for digital farming.
SMART AGRICULTURE
From environmental data to prediction.PROJECT OVERVIEW
AI + IoT
for Agriculture
SMART FASAL focused on developing digital infrastructure for wheat farming through cloud computing, Internet of Things and agricultural data.
Environmental observations collected from field devices could be captured, organized, visualized and analyzed through digital systems.
RESEARCH PIPELINE
Data to Prediction
DATASET 01
Waspmote Environmental Dataset
The documented Waspmote dataset contains environmental observations collected using six sensor variables during February to April 2021.
EXPERIMENT
Time-Series Forecasting
The documented experiment involved preparing the environmental time-series data, splitting training and testing datasets, tuning model parameters, building forecasting models and evaluating predictions.
Both ARIMA and LSTM approaches were explored for forecasting environmental variables.
RESEARCH FINDING
Variable Relationships
Soil Moisture Correlation
The analysis identified a high correlation between Soil Moisture 1 and Soil Moisture 2, approximately 0.99.
The documented analysis identified this as substantial redundancy and suggested approaches such as removing one variable or applying PCA when working with the LSTM model.
MY CONTRIBUTION