An elastica robot: Tip-control in tendon-actuated elastic arms

Abstract

Various applications ranging from pipeline inspection to robot-assisted surgery require highly maneuverable devices with precisely controllable tips. In this article, we propose and experimentally validate a mechanics-based approach to manipulate the quasistatic planar motion of the tip of a flexible arm by controlling the tensions in a pair of cables attached to its centerline. We show that by adopting a geometrically nonlinear elastica model for the arm, by carefully accounting for configuration-dependent tendon loadings, and by interpreting the problem of manipulating the tip of the arm as one of load optimization, it is possible to realize a high degree of accuracy. In the process, we also identify interesting features of the tendon-loaded elastica problem. Our tip-control strategy is based solely on a predictive elastica model, does not require any sensors and does not employ any feedback. The resulting elastica robot is slender, remotely actuated, energy-efficient and miniaturizable.