Date of Award
Doctor of Philosophy (PhD)
The work in this dissertation explores interfacing, modeling, and automation of continuum manipulators. In particular, we work to build a bridge between continuum robots and new areas of research through haptic interfacing, and explore how established and popular methods in the broader robotic community apply to continuum robots and their applications.
First, we describe a novel haptic interface designed specifically for the teleoperation of extensible continuum manipulators. The proposed device is based off of, and extends to the haptic domain, a kinematically similar input device for continuum manipulators called the MiniOct. The work in this dissertation describes the physical design of the new device, the method of creating impedance-type haptic feedback to users, and some of the requirements for implementing this device in a bilateral teleoperation scheme for continuum robots. We report a series of experiments to validate the operation of the system, including simulated and real-time conditions. The experimental results show that a user can identify the direction of planar obstacles from the feedback for both virtual and physical environments. Further, we discuss the challenges for providing feedback to an operator about the state of a teleoperated continuum manipulator.
Next, we explore methods for approximating continuum robot dynamics through rigid bodies connected together via compliant joints. The nature of the joints are designed to capture the material properties of continuum robot backbones, while the rigid links approximate the inertia and shape of the actual continuum robot's body. In evaluating the models, we formulate a state estimator based on energy minimization methods and compare the estimated state of a physical robot to ground truth values provided by external sensors. Finally we evaluate the model through a series of experiments testing the step response of the model and an experiment to observe the model throughout dynamic motion. The results show a good match between predicted motion of the simulated robot and the physical robot's state estimator.
Continuum robots have long held a great potential for applications in inspection of remote, hard-to-reach environments. In future environments such as the Deep Space Gateway, remote deployment of robotic solutions will require a high level of autonomy due to communication delays and unavailability of human crews. Further work in this dissertation explores the application of policy optimization methods through Actor-Critic gradient descent in order to optimize a continuum manipulator's search method for an unknown object. We show that we can deploy a continuum robot without prior knowledge of a goal object location and converge to a policy that finds the goal and can be reused in future deployments. We also show that the method can be quickly extended for multiple Degrees-of-Freedom and that we can restrict the policy with virtual and physical obstacles. These two scenarios are highlighted using a simulation environment with 15 and 135 unique states, respectively.
Finally, given the challenging nature of continuum robot modeling in comparison with traditional rigid-link robots, we consider the Kinematic-Model-Free (KMF) robot control method, previously shown to be extremely effective in permitting a rigid-link robot to learn approximations of local kinematics and dynamics (``kinodynamics'') at various points in the robot's task space. These approximations enable the robot to follow various trajectories and even adapt to changes in the robot's kinematic structure. In a final work in this dissertation, we present the adaptation of the KMF method to a three-section, nine degrees-of-freedom continuum manipulator for both planar and spatial task spaces. Using only an external 3D camera, we show that the KMF method allows the continuum robot to converge to various desired set points in the robot's task space, avoiding the complexities inherent in solving this problem using traditional inverse kinematics. The success of the method shows that a continuum robot can ``learn'' enough information from an external camera to reach and track desired points and trajectories, without needing knowledge of exact shape or position of the robot.
Frazelle, Chase Gilbert, "Continuum Robots: Interfacing, Modeling, and Automation" (2021). All Dissertations. 2827.