Building “First Expire, First Out” models to predict food losses at retail due to cold chain disruption in the last mile

Abstract

Current less-than-truckload (LTL) shipping practices allow for temperature abuse (TA) in the last segment (last mile) of the food supply chain. When this TA is combined with “First In, First Out” product rotation methods, it could lead to food spoilage and food waste; therefore, data-based decision models are needed to aid retail managers. An experiment was designed using pallets (4 layers/pallet × 5 boxes/layer) of commercially produced boneless chicken breast filet trays. The pallets were exposed to 24 h of simulated LTL TA (cyclic 2 h at 4°C, then 2 h at 23 ± 2°C). Filet temperatures were recorded for all 20 boxes using dataloggers with thermocouple wires. Additionally, microbiological sampling of filets [aerobic plate counts (APC) and psychrotrophic plate counts (PSY)] was conducted before (0 h of LTL TA) and after (24 h of LTL TA) the TA experiment for select boxes of the pallet and compared to control filets (maintained at 4°C). After TA, a shelf-life experiment was conducted by storing filets from predetermined boxes at 4°C until spoilage (7 log CFU/ml). Temperature and microbiological data were augmented using Monte Carlo simulations (MC) to build decision making models using two methods; (1) the risk of each box on the pallet reaching the bacterial “danger zone” (>4°C) was determined; and (2) the risk-of-loss (shelf-life < 4 days; minimum shelf-life required to prevent food waste) was determined. Temperature results indicated that boxes on the top and bottom layers reached 4°C faster than boxes comprising the middle layers while the perimeter boxes of each layer reached 4°C faster than centrally located boxes. Shelf-life results indicate simulated LTL TA reduced shelf-life by 2.25 and 1.5 days for APC and PSY, respectively. The first MC method showed the average risk of boxes reaching 4°C after 24 h of simulated LTL TA were 94.96%, 43.20%, 27.20%, and 75.12% for layers 1–4, respectively. The second MC method indicated that exposure at >4°C for 8 h results in a risk-of-loss of 43.8%. The findings indicate that LTL TA decreases shelf-life of chicken breast filets in a heterogenous manner according to location of boxes on the pallet. Therefore, predictive models are needed to make objective decisions so that a “First Expire, First Out” method can be implemented to reduce food wastes due to TA during the last mile.

Development of a predictive classification model and extraction of signature wavelengths for the identification of spoilage in chicken breast fillets during storage using Near Infrared Spectroscopy

“…2). Microbiological analysis of the stored chicken breast llets were performed using the protocol mentioned in Herron et al (2022). Different supervised ML regression models were also developed and compared for the prediction of logCFU/ml value.…”

“…Factors such as in-cell enzymes, nutrient availability, oxygen concentrations, and cross-contamination, along with temperature uctuations, contribute to food spoilage (Adegoke, 2004). For fresh poultry commodities, modi cations such as pH changes and fat oxidation can encourage micro ora growth and metabolism, leading to the formation of toxic compounds and off-avors (Sikorski et al, 2020;Huang et al, 2021;Herron et al, 2022).…”

“…Although fresh meat deterioration is a concern worldwide, more research is needed to understand the mechanisms and interactions causing spoilage (Herron et al, 2022). Several studies have investigated rapid detection techniques for spoilage in meat and related products.…”