Development of a freshness-assessment model for a real-time online monitoring system of packaged commercial milk in distribution. (May 2016)

• Record Type: Journal Article
• Authors: Kim, Byeong-Sam; Lee, Minwoo; Kim, Ji-Young; Jung, Jae-Yoon; Koo, Junemo
• Is Part Of: Lebensmittel-Wissenschaft + Technologie =. Volume 68(2016)
• Page Start: 532
• Page End: 540
• Publication Date: 2016-05
• Subjects: Predictive modeling -- Microbial growth -- Dairy -- Fluctuating temperature -- Shelf-life -- Spoilage -- Safety; Food industry and trade -- Periodicals; Food -- Composition -- Periodicals; Microbiology -- Periodicals; Nutrition -- Periodicals; 664.005
• DOI: 10.1016/j.lwt.2015.12.049
• Languages: English
• ISSNs: 0023-6438

Abstract

Abstract: The freshness of packaged pasteurized milk under fluctuating temperature condition in distribution, which was assessed by measuring the total microbial count in milk samples, was monitored by applying real–time temperature logging technology, which included RFID (Radio Frequency Identification), WSN (Wireless Sensor Networks) and predictive microbiology. The freshness assessment model, which would be used in a real-time online quality monitoring system for packaged commercial milk, was developed considering the impacts of large thermal capacity of packaged milk, wide biokinetic temperature range, i.e., 5–30 °C, and intermediate lags with temperature shifts as well as the choices of the primary and secondary models such as the Gompertz, the Roberts and Baranyi, and the Hills and Wright models. The justifiability of considering additional model parameters was statistically analyzed using Bayesian Information Criterion (BIC) to select the best freshness assessment model. The freshness assessment model with the Hills and Wright model as the primary model and the two-zone Arrhenius model as the secondary model was selected as the best one to use for the real-time milk freshness monitoring system. Highlights: A milk freshness assessment model was for online monitoring system. The first and second model constants were determined with full temperature histories. Statistical methods were used to compare models and select the best one. The intermediate lag was well captured by the new model.