Growth predictions of lettuce in hydroponic farm using autoregressive integrated moving average model

Abstract

Hydroponic farming techniques can grow plants faster when intensive monitoring is carried out. However, sometimes proper monitoring does not occur, which causes some plants not to grow as expected. This research uses a sensor to measure a parameter that affects the growth of the hydroponic plants and uses a camera to take pictures periodically to measure daily growth. The research presented in this article is to build a model and forecasting plant growth in hydroponics farming using a time series approach. This paper demonstrated that the historical data on lettuce growth could be used to predict future plant size. The autoregressive integrated moving average (ARIMA) model was used and analyzed according to the six performance criteria: mean absolute error (MAE), root mean squared error (RMSE), mean absolute percentage error (MAPE), Akaike information criterion (AIC), and Bayesian information criterion (BIC). To find the best model, different autoregressive (p) and moving average (q) parameters were examined. We find that the appropriate statistical model for lettuce growth is ARIMA (2, 2, 1) which has the lowest AIC, BIC, and MAPE with values of 76.67, 79.02, and 0.04, respectively, to forecast the plant size for the next three days.