Interface issues in robot scrub nurse design.

Abstract

The idea of a robot providing assistance in the medical field is very promising. Due to the myriad of unresolved issues a full therapist robot is still out of our reach. In this study we approach the problem in an application area that is sufficiently constrained to allow us a certain degree of practical success. Concretely we present a system description of a robotic scrub nurse (RSN) for microsurgery. We identify robot interfacing as a crucial problem to be solved. In that regard we examine how spoken language could be utilized, how vision could guide robot motion, and finally we examined human-robot interaction at the haptic level. For the haptic interface we have designed and evaluated a shape conforming grasp mechanism. Such an approach is uncommon in the robotic industry where per-task adapter-style exchangeable grasping mechanisms are mounted on the robot for each run. While industrial robots deal with few instruments and must lift heavy weights, our problem domain contains hundreds of instruments and prohibits any type of crushing or cutting movements. Our results provide an integration study of the various components for assistant robots. Furthermore we contribute novel insights in the design of grasp planning methods, visual tracking methods and natural language dialogue systems.