Online Kinodynamic Motion Planning
The aim of this project is to provide online and safe kinodynamic motion planning algorithms with completely unknown/uncertain dynamics, based on continuous-time Q-learning. We utilize integral reinforcement learning to develop tuning laws for the online approximation of the optimal cost and the optimal policy in continuous-time. We modify motion planning techniques to perform efficient online local re-planning by employing topological tools.
Keywords: Motion Planning, Reinforcement Learning, Optimal Control, Game Theory.
Anthropomorphic Robot Hands
This research project contributes to the development of a new class of anthropomorphic, adaptive robot hands. The robot hands can operate in unstructured environments and can achieve various grasping and in-hand manipulation actions, with only 4 actuators in a synergistic fashion. The design is optimized according to the human hand characteristics that are based on neuroscience, anthropometric studies, and medical tests. The fabrication procedure consists of 3D printing, and Hybrid Deposition Manufacturing (HDM) techniques.
Keywords: Adaptive Robot Hands, Compliant Mechanisms, Dexterous Manipulation.
This reserach project is focused on the design and development of an anthropomorphic, underactuated, personalized robot hand of low cost and weight. Elastic joints were selected in order to introduce passive compliance in the hand's structure, simplify the control problem and perform robust grasping in unstructured environments. We employed a single actuator with a novel selectively diﬀerential mechanism that allows the execution of diﬀerent grasping postures and gestures. The proposed design is parametric and allows the replication of personalized, anthropomorphic prosthetic hands. The fabrication procedure consists of either a laser cutting or a 3D printing technique.
Keywords: Prosthesis design, Underactuation, Differential Mechanisms.