Design of a collaborative architecture for human-robot assembly tasks

Collaborative robots, the so-called cobots, that work together with the human, are becoming more and more popular in the industrial world. An example of an application where these robots are useful is the assembly task. In this case, the human and the robot complement each other. On one side, the human can achieve more dexterous tasks, while on the other side, the robot can assist the assembly process to lower the physical and cognitive work load, e.g. to avoid errors, and in the same way reduce absenteeism. This paper describes a novel collaborative architecture for human-robot assembly tasks. The developed architecture is composed of four modules; face recognition, gesture recognition and human-like robot behavior modules are used to enhance the human-robot interaction, while the visual inspection module is utilized for quality control during the assembly process. A collaborative task consisting of the assembly of a box whereby the robot assists the human was designed and implemented on the Baxter robot. This was used as the application use case to validate the developed collaborative architecture.