To eat or not to Eat: Modeling food waste behavior at a hotel breakfast buffet

Food loss and waste are increasingly recognized as the outcomes of complex, interconnected social and environmental dynamics rather than isolated individual choices. In this study, we adopt a complex systems approach to explore how micro-level behaviors and interactions within commercial dining settings give rise to emergent patterns of food waste. We present an agent-based model that integrates a psychologically realistic representation of individual decision-making, grounded in the HUMAT socio-cognitive architecture. Here, artificial agents balance multiple motives — including social, experiential, and values-based drivers — while operating within a context shaped by the Motivation, Opportunities, and Abilities (MOA) framework from the social sciences. The simulated population represents diners in buffet-style environments. Individuals follow rules influenced by their motivations, opportunities, and abilities, which dictate factors such as the timing of their meal, portion sizes, frequency of servings, and amount of leftovers on their plates. Results highlight that the motives of conformism and sustainability have the strongest impact on food waste levels, with higher values drastically reducing leftovers. The model also shows that extra servings—particularly in scenarios with normal plate sizes—are strongly associated with increased waste. These findings underscore how behavioral drivers and situational constraints interact to shape food waste patterns, supporting the model's application in simulating context-sensitive interventions for promoting sustainable and responsible consumption.