No 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.

This paper explores a two state rover concept called the Transforming Roving-Rolling Explorer (TRREx). The first state allows the rover to travel like a conventional 6-wheeled rover. The second state is a sphere to permit faster descent of steep inclines. Performance of this concept is compared to a traditional rocker-bogie (RB) architecture using hi-fidelity simulations in Webots. Results show that for missions involving very rugged terrain, or a considerable amount of downhill travel, the TRREx outperforms the rocker-bogie. Locomotion of the TRREx system using a continuous shifting of the center of mass through “actuated rolling” is also explored. A dynamics model for a cylindrical representation of the rover is simulated to identify feasible configurations capable of generating and maintaining continuous rolling motion even on sandy terrain. Results show that in sufficiently benign terrain gradual inclines can be traversed with actuated rolling. This model allows for increased exploration of the problem's design space and assists in establishing parameters for an Earth prototype.

No abstract

No abstract