Assessing Reconfigurable Design for a Chaotic Objective in a Mars Exploration Rover

Abstract

This paper presents a strategy for assessing reconfigurability at the architecture level in the design of a complex system to fulfill a chaotic objective. A hi-fidelity simulation environment is used to quickly run a myriad of test scenarios on a reconfigurable system and a more traditional alternative. Specifically, the design of a mobility system for a Mars rover is used in this assessment. Rovers provide an excellent illustration of the issue; 1) the mobility problem is one with many random variables, 2) rover architectures are complex enough that a make-it, break-it, fix-it approach is prohibitively expensive and time consuming, and 3) several rover architectures employing reconfigurability have been proposed or are already under development. The ultimate objective of this research is to compare complex systems across a widely diverse range of architectures and reconfigurations. This paper is a step toward that goal. Four roverstwo architectures using two size scales-are tested in twenty terrain challenges, and their performance is explored across various levels of ground traction, slope, and rock field density. By considering potential missions as a combination of the objectives of the terrain challenges, architecture selection is explored. Performance is the primary measure considered in this paper with cost, risk, and other considerations to be addressed in future work. I.

Tradespace Exploration of Reconfigurability using Surrogate Points to Simplify Dominance Determination

“…7. Examples of the sparse (above) and dense (below) rock fields used in the rover testing 10 American Institute of Aeronautics and Astronautics…”

“…In previous work, the authors began the process of assessing rover performance against chaotic terrain 10 . This paper builds directly on this work so the highlights of that paper are presented here.…”

“…for more details. For each trial, a score was assigned for mean free path ratio (MFPR), root mean square distance from the path (Drms), and average speed in the trial (Vavg) 10 . Mean free path (MFP) is the average distance the rover can travel in a particular environment without encountering an insurmountable obstacle.…”

“…A challenge associated with this type of system is that analysis for these systems is non-trivial. Previous work by the authors explored this issue by studying the performance of a rover that could reconfigure when traversing chaotic terrain by transitioning between two main modes -rolling and driving 10 . This rover design was compared to a traditional rocker bogie design for a variety of terrain challenges.…”

Multi-modal reconfigurable robotic platform with 3D-immersive telepresence system

“…Reconfigurable robots have been used in demining, undersea experiments and planetary exploration missions, such as NASA's Mars rover Curiosity [3], to reduce the risk and eliminate the physical presence of humans from hazardous environments. However, absolute replacement of humans with robots is not yet attainable due to the limited level of intelligence of existing robotic systems.…”