An Architecture for Intuitive Programming and Robust Execution of Industrial Robot Programs

Abstract

Industrial robots are established tools in a variety of automation processes. An important goal in actual research is to transfer the advantages of these tools into environments in small and medium sized enterprises (SME). As tasks in SME environments change from time to time, it is necessary to enable nonexpert workers to program an industrial robot in an intuitive way. Furthermore, robot programs must concern unexpected situations as many SME environments are unstructured, because humans and robots share workspace and workpieces. The main contribution of our work is a robot programming architecture, that concerns both aspects: Intuitive robot programming and robust task execution. Our architecture enables users to create robot programs by guiding a robot kinesthetically through tasks. Relevant task information are extracted by an entity-actor based world model. From these information we encode the demonstrated task as a finite state machine (FSM). The FSM allows the reproduction and adaption of a task by the robot in similar situations. Furthermore, the FSM is utilized in a monitoring component, which identifies unexpected events during task execution.