A Configurable Skill Oriented Architecture Based on OPC UA

Over the last years, research done in automation and industrial robotics has established the foundations for skill-oriented systems based on the OPC UA standard. Nevertheless, utilizing these advances in other areas of robotics research can be challenging and time consuming. We present a framework aiming to reduce this entry threshold. Our solution is an open source, easy to configure tool based on OPC UA, that provides with a hardware agnostic, skill-oriented, event driven interface to systems. The framework allows integrating external hardware and software by means of plugins. It also provides a mechanism for endowing skills with hardware agnostic motion control. We demonstrate how our framework can be combined with state of the art approaches to control in a simulated assembly task with multiple robots, conveyors, actuators and sensors involved.

I. Introduction

Advances in intelligent robotics are enabling the de-ployment of flexible systems for different applications. In particular, the research done in the field of task-level pro-gramming using robot skills [1] has played an important role in this process. Task-level programming and skills allow reasoning about the behaviour of systems on a higher level of abstraction. As a result of this, some works have demonstrated how humans without expertise in robotics can take part in robot programming under this paradigm [2]. Also artificial intelligence can be combined with task-level programming for creating autonomous systems. Additionally, from a software point of view, robot skills can be combined in a modular manner, and naturally adopted in service oriented architectures (SOAs). This feature has made skills a key component for the deployment of agile manufacturing systems [3]. Hence, during the last years, new approaches in automation and industrial robotics are enabling a transition towards skill-oriented systems in manufacturing, that should eventually crystallize in new standards.