Högskolan i Skövde

In recent years human-robot collaboration has been an important topic in manufacturing industries. By introducing robots into the same working cell as humans, the advantages of both humans and robots can be utilized. A robot can handle heavy lifting, repetitive and high accuracy tasks while a human can handle tasks that require the flexibility of humans. If a worker is to collaborate with a robot it is important to have an intuitive way of communicating with the robot. Currently, the way of interacting with a robot is through a teaching pendant, where the robot is controlled using buttons or a joystick. However, speech and touch are two communication methods natural to humans, where speech recognition and haptic control technologies can be used to interpret these communication methods. These technologies have been heavily researched in several research areas, including human-robot interaction. However, research of combining these two technologies to achieve a more natural communication in industrial human-robot collaboration is limited. A demonstrator has thus been developed which includes both speech recognition and haptic control technologies to control a collaborative robot from Universal Robots. This demonstrator will function as an experimental platform to further research on how the speech recognition and haptic control can be used in human-robot collaboration. The demonstrator has proven that the two technologies can be integrated with a collaborative industrial robot, where the human and the robot collaborate to assemble a simple car model. The demonstrator has been used in public appearances and a pilot study, which have contributed in further improvements of the demonstrator. Further research will focus on making the communication more intuitive for the human and the demonstrator will be used as the platform for continued research.

Industrial robot manufacturers have in recent years developed collaborative robots and these gains more and more interest within the manufacturing industry. Collaborative robots ensure that humans and robots can work together without the robot being dangerous for the human. However, collaborative robots themselves are not enough to achieve collaboration between a human and a robot; collaboration is only possible if a proper communication between the human and the robot can be achieved. The aim of this paper is to identify and categorize technologies that can be used to enable such communication between a human and an industrial robot.

Human-robot collaboration (HRC) is the concept of combining a human and a robot into the same production cell and utilize the benefits of both. This concept has existed for more than a decade, but there are still quite few implementations of HRC within the manufacturing industry. One reason for this is the lack of knowledge when it comes to suitable tasks for HRC. Current research studies on the topic are mainly based on theoretical reasoning and/or research experiments, and little is known about what the industry perceive as suitable tasks for HRC. Therefore, this paper aims to investigate this and find out what industrial actors thinks are the most value-adding tasks for a human and a robot to carry out together. An in-depth interview study is undertaken with two companies and shop-floor operators, production engineers and automation engineers are interviewed. The result of the study pinpoints a number of tasks that the companies thinks are beneficial for HRC, which can serve as a guideline for other manufacturing companies considering to implement HRC.

This paper presents 'virtual collaborative robot', a virtual reality platform for designing and evaluating collaboration between operators and industrial robots. Within the platform, human-robot collaboration scenarios can be created and a user can interact with a robot without the safety risks that might arise with physical industrial robots. In an initial evaluation of the platform a scenario was implemented combining speech recognition, haptic control, and augmented reality to assemble a car model. The results from this evaluation indicate that the suggested platform can be used to successfully test new applications with the standard equipment of virtual reality headsets.