Novel design, stochastic motion planning, and control of robot systems supporting medicine

This dissertation focuses on robotics technology and its applications to medicine. Major research topics can be divided into two main categories: novel designs of robot systems for medical applications and probabilistic path/motion planning of robot systems in stochastic environment. We consider a special medical needle device and a small vibration robot for research on novel designs of robotic systems. An optimal insertion strategy for a flexible medical needle to minimize patients' tissue damage in 2D and 3D spaces, and a new hardware design to improve performance of the existing flexible needles which have a single and relatively low curvature in their trajectory are suggested. Additionally, a new mechanism of a small vibration robot design is studied for robot-assisted diagnosis procedure. This robot is propelled by vibration excitation and it provides the motivation for capsule endoscopes that can move around inside of the human body. We successfully developed a method to actuate the robot by the vibration excitation and then updated it to move on a 2D plane with a miniaturizable design. As to the optimal path/motion planning, probability-based path/motion planning methods are applied and improved for various robotic systems with a certain amount of noise. It is shown that an optimal path/motion that the robotic systems should follow to reach desired targets with the highest targeting probability can be obtained using the proposed planning method. Finally, the model parameters of actual robotic systems are estimated to improve the performance of the probabilistic path/motion planning. Those parameters can be estimated by comparing the analytic equations for statistical information and the experimentally sampled data. This dissertation contributes to improve medical procedures of special needles and robotic systems to minimize patients' damage, increase operation accuracy, and advance the reliability of robot technology in medical applications.