Empowering Worker-Robot Collaboration: Leveraging LLMs for Extracting and Visualizing Robot Task Metrics

In the context of Industry 5.0, which is characterized by a close integration between digital technology, industrial production, and human-centered design, collaborative robots emerge as key players. These robots are no longer isolated machines but an integral part of an interconnected ecosystem, where the fluidity of data plays a crucial role. Collaborative robots facilitate flexibility, efficiency, and safety in operations. However, this also introduces novel programming and data management challenges. A distinctive feature of collaborative robots is their ability to be programmed and used by non-expert users. This democratization of access to robotics offers significant advantages but also requires careful design of tools and interfaces to enable easy access to the data generated by the robots. In this context, the user interface assumes a pivotal role in ensuring that even those lacking programming expertise can fully benefit from the capabilities of collaborative robots and the data they produce. This study exploits Human Work Interaction Design principles to examine the problems encountered by individuals lacking programming skills when attempting to obtain data about robot performance. A solution that exploits Large Language Models is proposed.