| Summary: | Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Mechanical Engineering, 2001. === "June 2001." === Includes bibliographical references (leaves 66-71). === Exploratory space missions of the future will require robotic systems to lead the way by negotiating and mapping very rough terrain, collecting samples, performing science tasks, and constructing facilities. These robots will need to be adaptable and reconfigurable in order to achieve a wide variety of objectives. Conventional designs using gears, motors, bearings, encoders, and many discrete components will be too complex, heavy, and failure-prone to allow highly-reconfigurable systems to be feasible. This thesis develops new concepts that may potentially enable the design of self-transforming space explorers. The vision of this research is to integrate compliant bistable mechanisms with large numbers of binary-actuated embedded smart materials. Compliant mechanisms are lightweight and robust. Binary actuation is the idea of using an actuator in a discrete on/off manner rather than in a continuous manner. A binary actuator is easy to control and robust, and by using tens or hundreds of binary actuators, one can approximate a continuous system, much like a digital computer can approximate an analog system. The first part of this thesis examines the fundamental planning issues involved with systems having large numbers of binary actuators. The notion of a workspace is described and applied to the optimization of a manipulator design. Methods for solving the forward and inverse kinematics are discussed in the context of this application. These methods are extended to the trajectory and locomotion planning problems. Methods for planning systems of substantial complexity are developed in the context of exploratory space robotics. The second part of this thesis presents experimental demonstrations that examine elements of the concept. The results of several design prototypes are discussed. === by Matthew D. Lichter. === S.M.
|
|---|
