Caisim Growth Model Using Smart Farm IoTBased Machine Learning

Abstract

Abstract— Indonesia is a developing country where the majority of people still work as farmers. The caisim plant is one of the most extensively farmed plants. Caisim production increases year after year. Farmers, on the other hand, have been unable to satisfy the demands of a very high market demand because they frequently encounter crop failure due to caisim plants harmed by pests and diseases. The difficulties encountered are due to the scenario for remote plant monitoring and a lack of understanding about growth parameters in caisim plants. Based on these issues, the development of a website and growth prediction model will provide a solution for producing caisim plants with optimal growth. When measuring QoS (Quality of Service) for data transmission from the tool to the website, the average latency was 371,57 ms. During the QoS test, the average throughput for reading data from the device to the database was 3469,14 bit/s. Meanwhile, data for the plant growth prediction model was retrieved from the website and converted to a CSV file dataset. In this prediction model, the KNN approach was used. This algorithm will generate classification results in the form of optimal and nonoptimal values for each characteristic used.
Keywords — smart farm, caisim, MySQL, machine learning, IoT.