Visualization of Multi-Level Neural-Based Robotic Systems

Abstract

Autonomous biological systems are very complex in their nature. Their study, through both experimentation and computation, provides a means to understand the underlying mechanisms in living systems while inspiring the development of technological applications. Experimentation, consisting of data gathering, generates predictions to be validated by experimentation on artificial systems. Computational models provide the understanding for the underlying dynamics, and serve as basis for simulation and further experimentation. The work presented here involves analyzing how predictive models can be generated from biological systems and then be used to drive robotic experiments; and conversely, how can results from robotic experiments drive additional neuroethological data gathering. This process requires a variety of visualization techniques in modeling and simulation of increasingly complex systems.