Cable-driven robots are often considered an attractive solution for applications which require a low-inertia actuator capable of high accelerations, despite the added complexity of the system and increased challenges of the design. Previous research has developed various techniques for analyzing the workspace of a cable-driven robot and several methods of designing a cable routing have been demonstrated which enable a robot to perform a given task. In general, there are many choices of cable routing which satisfy the particular design requirements, and the problem is focused on how to select the best design from among the available choices.
Using techniques developed from statistical learning theory, the field of Randomized Algorithms offers explicit bounds on the estimation errors of an optimal solution, provided certain explicitly defined sample complexity requirements are satisfied. The main advantage of this approach is a quantitative understanding of the uncertainty of the result and the accuracy with respect to the true optimum to enable an informed decision regarding the sufficiency of the solution. The desired accuracy and confidence parameters for the optimal solution are tailored for the specific problem, allowing for balance between computational limitations and desire for solution quality. In this work, a summary of the randomized algorithm approach is presented, and an optimized solution to the design of a cable-driven manipulator is developed using a randomized algorithm according to specified accuracy, level and confidence parameters.