Analytical configuration of wheeled robotic locomotion

Through their ability to navigate and perform tasks in unstructured environments, robots have made their way into applications like farming, earth moving, waste clean-up and exploration. All mobile robots use locomotion that generates traction, negotiates terrain and carries payload. Well-designed robotic locomotion also stabilizes a robot's frame, smooths the motion of sensors and accommodates the deployment and manipulation of work tools. Because locomotion is the physical interface between a robot and its environment, it is the means by which it reacts to gravitational, inertial and work loads. Locomotion is the literal basis of a mobile robot's performance. Despite its significance, locomotion design and its implications to robotic function have not been addressed. In fact, with the exception of a handful of case studies, the issue of how to synthesize robotic locomotion configurations remains a topic of ad hoc speculation and is commonly pursued in a way that lacks rationalization. This thesis focuses on the configuration of wheeled robotic locomotion through the formulation and systematic evaluation of analytical expressions called configuration equations. These are mathematical functions which capture quantitative relationships among configuration parameters (e.g., wheel diameter, chassis articulation location), performance parameters (e.g. drawbar pull, maximum gradeable slope) and environmental/task parameters (e.g. soil geophysical properties, density and size of obstacles). Solutions to the configuration equations are obtained in parametric form to allow for comprehensive characterization of variant locomotion concepts as opposed to searching for point designs. Optimal configuration parameters are sought in the context of three indices of performance: trafficability, maneuverability and terrainability. The derivation of configuration equations, the estimation and optimization of configuration parameters and predictions of performance are performed in a computational framework called Locomotion Synthesis (LocSyn). LocSyn offers a practical approach to rationalizing configuration design of robotic locomotion through quantitative studies. The configuration of Nomad, a planetary prototype robot for exploration of barren terrain is a case illustrating the implementation and evaluation of the Locomotion Synthesis (LocSyn) framework put forth by this thesis.