Design and development of medical cobot to assist surgeon

Abstract

This paper presents an automated collaborative robot that works with a surgeon during a surgical operation incorporated with Machine learning algorithms and Deep learning algorithms to assist the surgeon by performing auxiliary actions. Our main objective is to create a framework for a collaborative nurse robot that upon voice instruction, will recognize and classify the surgical instruments present in the surgical tray and execute manual tasks such as picking up and delivering that instrument to its operator. There have been discussions on the types of co-bots and the types that can embrace the prescribed idea. After visualizing a co-bot model, designing it in CAD software is a crucial initiation to judge the practicality of the robot. The robot was designed to test its navigation before prototyping it. The main objectives to accomplish the working of an assistant medical co-bot is discussed. For object detection Aruco marker that uses Convolution Neural Network is implicated, for voice detection Google's speech recognition system having Deep neural network algorithm is embraced and for training the Co-bot using Arduino libraries is being discussed. Python has been used as a programming language and pyserial to communicate with Arduino from other two systems i.e., voice recognition system and object detection system. This medical Co-bot reflects our effort to lessen stress on surgeon and human error during operation.