Towards kinematic classification schemes for planetary surface locomotion systems

Abstract

As the planetary exploration programme matures, the objectives for robotic missions become more complex, and place greater demands upon the reliability and robustness of the autonomous systems created for these tasks. Consequently, a situation has arisen where the design of mechanical architectures may be insufficiently flexible to take advantage of the latest developments in control techniques. In particular, mechanical systems with sufficient functionality to tackle advanced autonomous operations, frequently exhibit inadequate robustness in the face of inevitable system faults to meet the requisite mission goals. This paper considers ways in which the kinematic structure of planetary exploration vehicles can be partly characterised using Graph Theory techniques. These include interchange graphs, their spanning trees, and their fundamental cycles and cutsets; adjacency matrices; degree sequences, and characteristic polynomials. It goes on to assess how such a characterisation may be used to define those kinematic features which make for successful, robust systems, well fitted to the complex mission goals required. Consideration is given to whether theoretical kinematic classification could provide a worthwhile building block in the development of novel design tools for future autonomous system development. A valid methodology for analysing and representing system characteristics is derived and discussed, and potential for further enhancement of the approach is identified.