Reducing perishable food loss through demand forecasting: empirical evidence from the meat industry

OPSEARCH · Springer Nature · 2026-11-19 · DOI: 10.1007/s10479-026-07456-8

Аннотация

Demand forecasting is a critical decision-making task, especially within the perishable food supply chain, where products exhibit highly variable demand patterns and limited shelf lives, often resulting in significant operational challenges as well as food loss and waste. Accurate forecasting is therefore essential for firms seeking to reduce food loss while improving operational efficiency performance and supply chain sustainability, with potential economic benefits. Using the real dataset of a leading company operating in the processed meat industry, this study develops and empirically evaluates a categorisation-based approach to demand forecasting, in which items are grouped by demand pattern and the most effective statistical or machine learning model is selected for each group. We evaluate the operational impact of improved forecasts through a simulation-based analysis on inventory-related indicators, measuring food loss reduction as well as warehouse inventory levels and dynamic production adjustments. Results show significant improvements in the operational and environmental performance of the company, including a 27% average food loss reduction compared to the current state.

Читать у издателя →

Похожие статьи