Prediction of Acidity Levels of Fresh Roasted Coffees Using E-nose and Artificial Neural Network

摘要

As for the coming automation age, development of the sense of robot including sight, smelling, hearing and touch is vital for robots to complement complex tasks of human. Recently, increasing interests in robotic chef and barista call for the development of “digital deliciousness” technology, in order for the robots to have a capability of food tasting. This work demonstrates a preliminary development of gourmet robot by using electronic nose (e-nose) technology to determine the scoring and cupping of the quality of coffees as compared to the human testers. In this study, it was mainly focused on the acidity levels of fresh roasted coffee. Array of different eight semiconductor gas sensors was used to smell the coffee's aroma. Electronic nose can clearly classify the acidity levels of different roasting degrees of the roasted coffee and has nearly the same results of scoring as obtained by using artificial neural network and the human's scoring. Thus, the e-nose has shown its capability for integration into the gourmet robot according to this study.