Evaluation of a Hypervisor-Based Smart Controller for Industry 4.0 Functions in Manufacturing

In machine tools, predictive maintenance and process monitoring can help optimize the efficiency of production systems by increasing machine availability, product quality, and by early fault detection. When integrating new sensors to provide data for such Industry 4.0 functions, the required interfaces and processing capacity are not always available, especially when upgrading existing production systems. An approach to face these issues is the smart controller architecture for sensor integration and data pre-processing. It uses a hypervisor to enable both fast reaction to incoming data and the flexibility and driver support of a Linux environment on a single platform. In this work, we implement and evaluate this concept on a commercial off-the-shelf (COTS) single-board computer in a real-world application. We show how the platform can be used for the integration of a state-of-the-art spindle monitoring sensor with an industrial edge device. Finally, we investigate the achievable end-to-end latency using the smart controller concept in comparison to a native Linux setup both in general and with respect to the application.