Timing-decisions by smell for shaking and firing extremely affect partially oxidized tea quality. A complete processing of Bao-Chung or Oolong tea takes 14 to 18 hours. It is hard to highly focus always. Therefore, assisting system to accurate determine the smell will be critical for tea quality and precision tea production. To solve the problem, we designed an electronic nose sensing array for aroma molecules during tea processing. Through the algorithm, we evaluated the representative aroma molecules by monitoring scents. Using the established e-nose system, big data from modulating-smell experiments developed a database which is based on aromatic profile related to tea quality. This intelligent tea manufacturing system is capable to real-time monitor, and the instructions of "shaking" and "firing" will be announced automatically. After in-site testing, the tea produced by the "intelligent tea-manufacturing system" can achieve the effect of making consistent high-quality tea.

Searching for key aromatic molecules is critical for the application of electronic nose. This technique proposes the key scents during partially oxidized tea processing, and develops the corresponding electronic nose system. We establish a tea processing logic based on the smell, which is composed by aromatic molecules, by electronic nose.According to the relationship of aroma and tea processing, the first intelligent tea processing technique which is capable of monitoring tea aroma and giving process instructions in the tea manufacture process. This technique solves the past problem that the tea master can only rely on the highly sensitive sense of smell to make process decisions.

The manufacturing process of partially oxidized tea in Taiwan is outstanding of the world. However, the tea making is only relied on the sensitive sense of smell and the experience by tea masters. The limiting factor is mainly on the lack of a key aromatic profile for advising system of tea processing. Our "intelligent tea manufacturing system" is able to evaluate the content of key aroma molecules digitally, and using the intelligent aroma calculating system to perform automated manufacturing process advises. This technique can be connected in series with automated equipment to achieve the aims of industrialized manufacturing of high-quality partially oxidized tea.