Computer systems integrate motivational properties to enable autonomous, proactive, and goal-directed behaviors, inspired by theories of human motivation. These properties are relevant both for autonomous agents and for systems that simulate human behavior. However, many existing models overlook the kinetics and interactions among physiological variables, which limits the accurate representation of internal needs and the adaptability of behavior over time. This paper proposes a computational model for generating physiological needs —specifically thirst— by monitoring bloodstream biomarkers such as sodium concentration and osmolarity. The model highlights the dynamic interactions of these biomarkers to trigger motivational states that lead to goal-directed behaviors, such as fluid intake or exploratory actions. The model’s validity is assessed by comparing its outcomes with findings from neuroscientific studies on thirst regulation.