Technological advancements in agriculture are continuously evolving to meet the growing demand for food. In recent years, agricultural technology has been shifting towards a new paradigm: Agriculture 4.0. However, a new role within this system needs to be introduced. Various rover models have been developed in recent years to enable autonomous decision-making. This thesis focuses on studying the dynamic behavior of such vehicles in an environment that simulates real-world conditions, proposing improvements. The first phase explores the steering systems commonly adopted in agriculture, followed by a brief theoretical explanation of their operation, evaluating which system might yield better results theoretically. Subsequently, multi-body simulations are conducted to study the dynamic behavior, reporting contacts and a structure as close as possible to the real rover.