The development of a discrete-time display quantity optimization algorithm for reducing food loss and increasing profits in retail stores

Abstract

In recent years, the significance of addressing food loss has increased due to its environmental impact, economic loss, and concerns about the sustainability of food supply. This is especially relevant in retail stores, where large amounts of food loss occur, and its reduction is highly sought after. In this study, we developed a discrete-time display quantity optimization algorithm based on demand forecasting to tackle these challenges. This algorithm learns demand trends from past sales data and suggests the optimal display quantity to minimize food loss from excessive procurement and loss of sales opportunities. The optimal display quantity is output as a single value per unit time, eliminating the need to track detailed display status during operation. This makes the algorithm highly practical. In addition, we formulated optimization requirements and constraints in store operations and validated the effectiveness of the algorithm through numerical experiments and trial applications in actual store operations.