In this paper, we describe a system for generating three-dimensional visual\nsimulations of natural language motion expressions. We use a rich formal model\nof events and their participants to generate simulations that satisfy the\nminimal constraints entailed by the associated utterance, relying on semantic\nknowledge of physical objects and motion events. This paper outlines technical\nconsiderations and discusses implementing the aforementioned semantic models\ninto such a system.\n