Largest Realistic Environment Simulation for Robotic Control Software

Saturday, October 29, 2011
Hall 1-2 (San Jose Convention Center)
Jessica De Silva , California State University, Stanislaus, Hilmar, CA
Nicola Basilico, PhD , University of California, Merced, Merced, CA
Stefano Carpin, PhD , University of California, Merced, Merced, CA

Realistic environment simulation is instrumental for the development of robotic control software. The task requires testing the software in the simulation prior to its exposure in the real world to ensure proper operating. Therefore, the objective of this project was to create the largest and most realistic environment model to be used in experiments done by scientists. 3D models, known as static meshes, were designed using the 3DS Max program which, along with Maya, is the primary program used for blockbuster animations. These models are then imported into the newest and most efficient game engine, Unreal Development Kit, also known as UDK. This engine contains accurate rigid body simulation and can withstand the input of tens of robots within one environment. The environments simulated within UDK entails maximizing retention to detail to guarantee accuracy in the training of the robotic control software. With UDK’s monthly updated software, the purpose is to create simulations that are indecipherable among the real world.