Practical Comparison of Plant Pest and Disease Control Technologies Based on Neural Networks, IoT, and AI: A Systematic Review

Nowadays, coffee lovers are growing extensively. It can be seen from coffee's popularity as the second largest commodity in the world. Producing the best quality coffee beans is not an easy thing. Diseases and pests can attack a plant randomly. For example, coffee leaf rust disease causes a decrease in coffee production by up to 35%. Rust fungus on coffee plants can spread through the distribution of fungal spores by natural forces. Therefore, to increase the production of quality coffee beans, it is necessary to have technological support to help detect pests or diseases in coffee plants. Technologies that can be used include neural networks, the Internet of Things, and artificial intelligence in collaboration with deep learning. This article reviews and compares the three technologies researchers have experimented on. The Artificial Intelligence-Deep Learning (AI-DL) technology experiment conducted by Livio et al. produces an accuracy of 95%. Meanwhile, the experiments with IoT technology and neural networks conducted by Escola et al., Divyashri et al., and De Vita et al. resulted in an accuracy of 100%, 88.35%, and 96%, respectively. Even though high accuracy has been obtained, these results show that different modeling techniques can make a difference in accuracy in detecting disease from a plant.