Prediction of Tempe Raw Material Needs for Home Industry Using Tsukamoto Fuzzy Logic Algorithm

Abstract

The home tempeh industry is faced with the challenge of accurately predicting raw material needs. Implementing an effective prediction system can help estimate the amount of tempeh production in the future, so as to avoid stock shortages of raw materials. In addition, when tempeh producers have accurate predictions, they can manage soybean stocks better, especially when there is a stock shortage, so that production is not affected by rising soybean prices. This research aims to develop a Tsukamoto Fuzzy Logic based prediction system to overcome this problem. Research methods include collecting historical production data, designing fuzzy models, and system testing. Interim results show that the developed system is able to increase prediction accuracy by up to 95%, which in turn can reduce production costs and minimize waste of raw materials.