Abstract
This paper describes the implementation of inline near infrared spectroscopy measurements to determine the key quality attributes of caramel in an industrial manufacturing process. The current control strategy is based around offline sampling and laboratory analysis which is costly, time-consuming and compromises operational control. Moving to inline analysis overcomes these limitations but the development of robust calibration models is a challenging task. Product recipe changes and natural raw material variation compound the issues arising in typical calibration model development. It is demonstrated that with appropriate spectroscopic signal pre-treatment and pattern recognition algorithm analysis, it is possible to provide the operators with a robust assessment of the caramel key quality measures so that they can respond in a timely manner if further processing actions are required. Results demonstrate the ability for NIR measurement to detect batch deviations that subsequently were judged to fall outside of the quality control limits as determined in pre-release laboratory assessment.
| Original language | English |
|---|---|
| Article number | 111717 |
| Journal | Journal of Food Engineering |
| Volume | 361 |
| Early online date | 7 Sept 2023 |
| DOIs | |
| Publication status | E-pub ahead of print - 7 Sept 2023 |
Bibliographical note
Funding Information:This work was supported by a Knowledge Transfer Partnerships (KTP) program [partnership number 11523] between British Bakels Limited and Newcastle University in collaboration with the National Horizons Centre at Teesside University . The partnership is part-supported by a grant from Innovate UK and part-supported by British Bakels Limited .
Funding Information:
This work was supported by a Knowledge Transfer Partnerships (KTP) program [partnership number 11523] between British Bakels Limited and Newcastle University in collaboration with the National Horizons Centre at Teesside University. The partnership is part-supported by a grant from Innovate UK and part-supported by British Bakels Limited. We would like to especially express our thanks to the colleagues at British Bakels and to Andrew Leather from Stratos Control Systems for their support and contribution to the success of this project.
Publisher Copyright:
© 2023 The Authors