Physically Based Modeling Simulation Testbed For Robotic Exploration Of A Martian Surface


A rover's performance on extraterrestrial terrain can be predicted by simulating the physical components of the rover, its environment, and all of the physical interactions in the rover-terrain system. Existing computer simulation software has two characteristics that can be improved upon. First, most simulations treat terrain as a solid geometric surface and use theoretical physics to predict interactions with a rover. However, since soil or sand is usually a semi-fluid solid, this modeling technique disregards some of the important material properties of the terrain. Second, simulations that do use physically based modeling with finite element methods are very computationally intensive and require uncommonly powerful computers. The goal of this project is to create a physically based modeling software that runs on standard personal computers which can accurately simulate the interaction of Mars rovers with the Martian terrain and predict the performance of various rover designs on different terrain conditions. The simulation allows the user to set parameters such as the slope, friction, and other material properties of the surface, and give the rover initial conditions such as total available power and rover starting position. When running the simulation, the program can instantaneously extract rover performance data and output the data to a text file. The performance parameters of concern include the rover's instantaneous position, velocity, and wheel slippage. The user has the option to choose from two methods to simulate the Martian soil. The first is the standard method that most existing simulation software uses, which treats the ground as a large solid surface. The second method represents soil as a bed of line segments that can be deformed according to the interactions with the rover. Due to the semi-fluid nature of soil, the standard modeling technique is suspected to have difficulty in describing the motion of the soil as caused by the force of the rover's wheels exerted on it, thereby creating inaccuracies in the model. In this study, we are particularly interested in determining whether this dynamic modeling technique gives a more accurate simulation of real soil than the standard method. Copyright © 2005 by the American Institute of Aeronautics and Astronautics, Inc. All rights reserved.

Publication Date


Publication Title

A Collection of Technical Papers - 1st Space Exploration Conference: Continuing the Voyage of Discovery



Number of Pages


Document Type

Article; Proceedings Paper

Personal Identifier


Socpus ID

28744432231 (Scopus)

Source API URL


This document is currently not available here.