A Conceptual Framework for the Improvement of Robotic System Reliability Through Industry 4.0

In recent years, the traditional manufacturing industry has been affected by the advent of Industry 4.0. Robotic systems have become a standard tool in modern manufacturing, due to their unique characteristics, such as repeatability, precision, speed, and high payload. However, robotics manipulators suffer from low reliability. Low reliability increases the probability of disruption in manufacturing processes, maximizing in this way the downtime and the maintenance costs. Consequently, for complex systems, like robots which consist of a lot of interdependent components, with various failures, it is critical to develop effective reliability assessment techniques to ensure high performance. Until now, there are several reliability assessment methods, but they are mostly time-consuming and expert-knowledge-intensive processes. This research develops a model-based approach for the reliability calculation of a robotic cell pinpointing the simplifications that are taking into account. Furthermore, a data-driven approach is proposed with the aim to enhance the results of the model-based approach and to improve the reliability of the robotic system towards the new directions that Industry 4.0 technologies can offer, by predicting the Remaining Useful Life (RUL) of critical components, using real-time data, thanks to the Digital Twin (DT).