Enabling Robotic Adaptive Behaviour Capabilities for New Industry 4.0 Automated Quality Inspection Paradigms

摘要

The seamless integration of industrial robotic arms with server computers, sensors and actuators can revolutionize the way automated Non-Destructive Testing (NDT) is performed and conceived. Achieving effective integration and the full potential of robotic systems presents significant challenges, since robots, sensors and end-effector tools are often not necessarily designed to be put together and form a holistic system. This paper presents recent breakthroughs, opening up new scenarios for the inspection of product quality in advanced manufacturing. Many years of research have brought to software platforms the ability to integrate external data acquisition instrumentation with industrial robots for improving the inspection speed, accuracy and repeatability of NDT. Robotic manipulators have typically been operated by predefined tool-paths generated through off-line path-planning software applications. Recent developments pave the way to data-driven autonomous robotic inspections, enabling real-time path planning and adaptive control. This paper presents a toolbox with highly efficient algorithms and software functions, developed to be used through high-level programming languages (e.g. MATLAB, LabVIEW, Python) and/or integrated with low-level languages (e.g. C#, C++) applications. The use of the toolbox can speed-up the development and the robust integration of new robotic NDT systems with real-time adaptive capabilities and is compatible with all 6-DOF KUKA robots, which are equipped with Robot Sensor Interface (RSI) software add-on. The paper describes the architecture of the toolbox and shows two application examples, where performance results are provided. The concepts described in the paper are aligned with the emerging Industry 4.0 paradigms and have wider applicability beyond NDT.