Maize Plant Conditions Prediction Using IoT Systems and Machine Learning Techniques for Precision Agriculture

Maize is the major source of income in Africa, especially in Rwanda. However, there are many diseases which affect its growth as well as lowering its production. Using technologies like Artificial Intelligence (AI), the Internet of Things avoids the manual tasks which may come up with errors and help farmers to get automation of farms and the control of farm resources such as soil parameters, pests, and insects. This work highlights the IoT systems and Machine Learning Techniques applications in precision agriculture. The proposed solution uses NPK sensors to sense the soil quality/chemical properties, temperature, moisture, humidity, and EfficientNet deep learning model was used to predict maize plant healthiness. The output results shows that the model provides the best performance and can achieved 95% accuracy, and it can be seen that our model has reduced the loss from 79% to the 17%.