ViRTNC – Real-Time Virtualization of NC Controllers for Dynamic Provision of Hardware Resources for Compute-Intensive Control Tasks

Due to the demand for individual, high-quality and low-cost products in ever shorter time intervals, manufacturing companies are facing major challenges. Scarce resources and increasing sustainability demands further complicate the situation. To ensure their competitiveness, companies must work more productively and rely on increasing the productivity and accuracy of production and machine tools. To meet these demands, innovative control algorithms are being developed at the Laboratory for Machine Tools and Production Engineering at RWTH Aachen University, which, for example, make it possible to increase the dynamics of feed axes while maintaining quality. However, manufacturing companies are often unable to use such innovative control algorithms at a reasonable cost. Either powerful control hardware must be provisioned unnecessarily in times of low quality and productivity requirements, controls must be fundamentally oversized, or the potential of computationally intensive control software remains completely unused.

A potential solution to this problem is being investigated in the research project ViRTNC. The decoupling of control hardware and software by a virtualization layer should enable a dynamic scaling of computing power depending on the complexity of the control functions. Control software should thus be able to migrate seamlessly from one hardware platform to another. To exploit economies of scale, the possibility of consolidating multiple control systems on one edge server will also be investigated. The use of rigid, monolithic, and categorically oversized control systems, which in most cases are not utilized but are no longer adequate for emerging requirements, will thus be specifically avoided. Important considered topics for the realization of a corresponding automation system are real-time virtualization on commodity servers, real-time communication in Ethernet networks, and the management and orchestration of the resulting virtualized control systems.