A Robust Optimization Model for Determining Optimal Diets for Food and Beverage (F&B) Items

This paper presents a simple and easy-to-use methodology for designing the menu in Food and Beverage (F&B) enterprises over a period of time, considering that certain elements of the problem are subject to uncertainty. The methodology considers both nutritional and financial aspects and allows the decision makers to explore the effect of the uncertainty on the final solutions, according to their perception of risk.The proposed methodology is based on multi-objective mixed integer programming and in particular the Almost Robust Optimization (ARO) approach introduced by Baron et al. (2019). In contrast to conventional Robust Optimization techniques, the ARO approach is more flexible and offers the decision makers the possibility to express their attitude towards risk through appropriate parameters and obtain a series of solutions corresponding to different levels of risk.The proposed model is applied in a case study concerning F&B enterprises from the island of Crete, Greece, using real data that was collected in collaboration with nutritionists and managers employed in F&B enterprises.Decision-makers in F&B enterprises may use the proposed model as a decision support tool to incorporate the inherent uncertainty into the decision-making process. Through appropriate parameters, they may select optimal diets that are feasible for most realizations of the uncertain parameters, without incurring significant increases in cost. The model is flexible and produces a series of alternative solutions based on the decision makers’ preferences and perception of risk.