Abstract
This study reported an original end-to-end dataflow engineering framework for the quality transfer principle to overcome the quality challenges in real-world honey manufacturing. Firstly, 650 pivotal data points of physical and chemical quality attributes from 65 batches of honey intermediates were characterized through multiple sensors, which included rheological properties, acidity, moisture, and sugars. Furthermore, a hypersensitized TAS1R2@AuNPs/SPCE biosensor was developed to identify biological quality attributes of honey, the powerful affinities between honey intermediates and the TAS1R2 receptor were discovered (K(D) < 1 × 10(-8) M), and the abnormal batches of B2, B23 and C23 were diagnosed by TAS1R2@AuNPs/SPCE biosensor and multivariable algorithm. Finally, the end-to-end dataflow containing physical, chemical and biological critical quality attributes was successfully established to interpret the quality transfer principle of honey manufacturing, which revealed that the front-end refining process was relatively unstable and the back-end refining process was a negligible influence on the quality of honey manufacturing. This framework embraces quality management, quality transfer, and biosensor information, which will contribute to discovering the quality transfer principle in industrial innovation for intelligent manufacturing.