Carnegie Mellon Robotics Institute
My broad research interests span the spectrum from theoretical kinematics to applied medical and snake robotics. Starting with my PhD work at the Technion, Israel Institute of Technology, my research focused on the kinematics of parallel mechanisms and how they could be applied to medical applications, by assisting surgeons during operations in general, and in spinal operations in particular. Upon graduation, I came to Carnegie Mellon where I have continued my work in medical robotics with ICAOS lab at Western Penn Hospital, and initiated an effort with Howie Choset in snake robotics. I have also developed some new fundamental theory in grasping and multi fingered robotic hand. I took part of developing low level and high level controls for hyper-redundant robots.
My work in kinematics begins with my Ph.D. work in developing kinematics and analysis tools for parallel robots which are based on screw theory, line geometry and projective geometry theories. These kinematic tools determine whether a robot is in or near a singular configuration and the instantaneous motion the robot tends to perform, while in such a singular configuration. Moreover, these methods provide physical interpretations to the mechanism of singular configurations, hence allow prediction of singular configurations for a given robotic structure. These methods also provide a quantitative and qualitative measurement tool for the rigidity of a robot at each instantaneous configuration within its workspace.
Next, my work also provides a synthesis tool to design and construct parallel mechanisms. This tool optimizes the structure of the robot in order to maximize (or minimize) the instantaneous work generated by the robots structure to the robots end-effector while the effector is undergoing an instantaneous movement characterized by the task. Minimization of the instantaneous work or maximizing it is a byproduct of the robot task. The process can be used in the design phase to determine the robot's kinematic parameter as well as once the robot is deployed, to determine optimal configurations given the kinematic parameters. This method could also be applied to other mechanisms and structures. For example, it could be used to determine the configuration of a hyper-redundant robots (snake robots), and solve the structure redundancy in order to determine the instantaneous work generated by the robot's end-effector on the environment.
The kinematic tools which are presented above were used for the analysis and synthesis of a new, miniature, medical-robot for spinal operations, which is based on a parallel robotic structure. In order to design such a robotic system a morphmetric study of the human lumbar spine was conducted. The investigation fills the missing data in the literature regarding exact dimensions of the spinal anatomies. One of the consequents of this study is a definition of the required workspace from a medical robotic system during spinal procedure. In this study the entry points and angles for several pre-selected anatomies, in the vertebra, are calculated. Next we defined other design parameters of the robot, namely, the forces and moments applied by the physician during insertion of Kirchner wires to soft tissues and drilling in hard tissues. These forces and moments were measured during clinical experiments. Moreover, a theoretical model of the expected error of the robotic system due to bone deflection and mechanical structure deflection is derived, and examined clinically in the operating room. Finally, based on the results, a miniature parallel robot was constructed and is currently in stages of getting FDA approvals.
|Research Interest Keywords|
|biomechanics, kinematics, medical robotics, search and rescue|
|The Robotics Institute is part of the School of Computer Science, Carnegie Mellon University.|
Contact Us | Update Instructions