Human-in-the-loop: Modeling and Simulation

Agent-based simulation forms an innovative simulation paradigm with great potential. Such a model is based on the metaphor of a multi-agent system which can be shortly characterized as a system consisting of interacting actors in an also modeled environment. It is per se especially apt for people without extensive experience in math or formal methods, such as empirically working scientists, teachers or students in school or high schools. However, up to now there are still too many conceptual and technical challenges for people with such backgrounds. Problems can be found on both the actual modeling procedure for concretizing theories, as well as when demonstrating how complex phenomena are generated from interacting small-grain entities. This project aims at building a framework that systematically is adapted to the human user. This is managed by creating a combination of high-level languages for modeling of agent behavior in spatial environments as well as for animating complex agent activities. It shall be possible to represent agent behavior using language primitives such as "move around the obstacles, leaving them on your right side" instead of calculating convex hulls and paths around a set of obstacle polygones. Agent actions shall be connected to animation patterns so that directly from the running simulations nice 3d visualization can be generated without referring to low level 3d visualization.