Abstract:
The research titled “Harnessing Embedded Technology For Smart Tobacco
Agriculture With Arbuscular Mycorrhizal Fungi and Organic Fertilizers” show the
important insights about embedded technology based Smart Agriculture for Tobacco
cultivation. Tobacco cultivation need high level nutrients for its plant growth and also
tobacco cultivation on land create soil degradation problem. As a result, we can’t
cultivate tobacco on same land again and again. To mitigate this problem our finding’s
is that we use Arbuscular Mycorrhizal Fungi (AMF) to uptake important soil nutrients
like phosphorus, Potassium etc. Using 3 pin Npk sensor we collect Nitrogen,
phosphorus and potassium level on soil. Based on Phosphorus optimal value, AMF
mixed soil has appropriate phosphorus level as compare to Non-AMF mixed soil.
Gathering data from two type of data sources we labeled AMF mixed soil as
“Satisfactory” and Non-AMF mixed soil as “Unsatisfactory”. Our dataset consists of
8314 datapoints. After Preprocess our dataset we apply five machine learning
algorithm which are as follows: Gradient Boosting, Random Forest, SVM, DecisionTree, and Naïve Bayes. Among all models SVM and Decision has lowest accuracy
85%. Gradient Boosting outperformed all other model with accuracy 87% which was
highest accuracy. Our work will help farmers to classify their soil quality which was
suitable for their tobacco cultivation. This will reduce the use of Chemical fertilizers
on soil.
