Navigating in rough terrain is a complex task that requires the robot to be considered as a holistic system. Algorithms, which don't consider the physical dimensions and capabilities of the mobile robot lead to inefficient motion and suffer from a lack of robustness. A physical model of the robot is necessary for trajectory control. In this paper, quasi-static modeling of a six-wheeled robot with a passive suspension mechanism is presented together with a method for selecting the optimal torques considering the system constraints: maximal and minimal torques, positive normal forces. The aim of this method is to limit wheel slip and to improve climbing capabilities. The modeling and the optimization are applied to the shrimp rover.

Human testing of bicycle equipment can be subjective. It is often more desirable to establish objective measures of a system's performance. A bicycle transmission must withstand varying loads placed upon it by the rider and the riding conditions. It must also reliably shift between speeds through numerous cycles. This paper describes the design of a robotic rig for testing bicycle transmissions. The test rig simulates the pedaling and shifting of a human and allows for programmed resistance to simulate riding conditions. The test rig can be programmed to perform numerous shifting sequences under varied operating conditions for tests lasting several hours. Shifts can be performed every 1.5 seconds to accelerate the lifecycle of the transmission. This paper describes each of the systems of the test rig and highlights the design considerations in its development.