HUMAN IN THE LOOP CONTROL IN ROBOTICS FOR MANUFACTURING

Abstract

Traditional manufacturing processes are limited to either being fully automated (CNC machining, robotic packaging, robotic welding systems, etc.), or fully manual (assembly tasks, hand tool operation, etc.), with the automated processes being separated via safety barriers in work cells from the manual processes; however current collaborative robot systems are greying the divide in how human workers and machinery are separated in factories, by offering lower- risk force compliant systems which can reduce or eliminate the requirement for bulky and restrictive guarding. This progress toward guard-less machinery which can operate directly next to human workers opens up not only new ways in which technology can assist human workers; but also how human workers can assist robots. Although collaborative robotics opens up new environments to operate in, many issues will remain which prevent the use of robots for new tasks, instead of human workers, due to a human’s cognitive capabilities. Human-in-the-loop control systems may present a way for robots to expand their task capabilities by off-loading some of the cognitive processing to a human co-worker, forming a Human-Robot team which can perform greater than either used alone. This paper presents the results of early-stage testing of a human-in-the-loop system in which human participants controlled a simulated robot to accomplish a list of tasks. Manual and semi-autonomous control schemes were tested, where time to completion and number of collisions were recorded to measure the effectiveness of human-in-the-loop control over the fully manual system.