1. Why develop Hypervisor technology?
Traditionally, market consulting firms define and analyze markets based on products. PLCs and industrial PCs were divided into two distinct markets. Traditional PLCs do not run operating systems but execute commands directly through the hardware, which limits their functionality. Therefore, subsequent development focused on controllers based on Real-Time Operating Systems (RTOS). These operating systems excel at handling real-time control tasks, such as high-speed I/O sampling and high-speed response control tasks like motion control. General-Purpose Operating Systems (GPOS), such as Windows and Linux, typically lack real-time capabilities but are strong in graphics, simulation software, and visualization design. As shown in Figure 1, the Hypervisor integrates the capabilities of RTOS and GPOS.
Figure 1. Seamless integration of Hypervisor with RTOS and GPOS
Figure 2. Internal architecture of B&R Hypervisor
In the past, many users often configured multiple devices: one PLC for real-time tasks, one PC for non-real-time tasks such as HMI display (e.g., 3D animation, dynamically calculated trend charts), and sometimes even an additional PC for handling vision tasks alone. With the further development of chips towards multi-core architectures and the increasing urgency of integrated processing, the Hypervisor, as a more convenient solution, has been adopted by leading automation manufacturers like B&R for use in new control and computing architecture designs.
2. Advantages of B&R Hypervisor
B&R's Hypervisor typically runs on industrial PCs, such as the APC series industrial control computers, which usually use multi-core processors such as Intel Core-i series, Atom Apollo Lake processors, or Panel PC series products. The Hypervisor is also configured in Automation Studio, and real-time tasks are assigned to B&R's Automation Runtime, while storage, graphics computing, and other tasks are assigned to Windows/Linux systems, as shown in Figure 2. B&R's Hypervisor supports Automation Runtime Embedded systems and open general operating systems such as Windows/Linux.
It has the following significant advantages:
• Seamless integration: Achieve seamless integration of real-time and non-real-time tasks within a single hardware and software architecture;
• Real-time capability: Real-time tasks and Windows/Linux tasks communicate across multiple cores for extremely fast response;
• Stable and reliable: Because the hypervisor is embedded in the CPU core, its stability and reliability surpass those of traditional software-based systems;
• Cost advantage: Integrating the controller, HMI, and industrial PC into a single device reduces system hardware costs.
3 Typical Application Scenarios – Edge Computing
With the rise of smart manufacturing and industrial IoT applications, edge computing architecture has become a hot topic in industry. Edge computing, building upon traditional control tasks, enables devices to perform more global optimization, scheduling, and strategic tasks through data connectivity. These tasks differ from signal-based control; they are based on the integration of more information. Their data types and required task processing are better suited to architectures like Windows/Linux, including highly dynamic tasks such as machine learning and local intelligent inference. Simultaneously, the computation results need to be dynamically fed back to the RTOS.
Figure 3 A typical application scenario
Hypervisor is used to command the operation of robots, motors, hydraulics and other actuators. Therefore, a local edge computing architecture can be built using Hypervisor, as shown in Figure 3, which is a typical edge computing architecture.
In edge computing scenarios, it encompasses numerous application requirements:
• Large-capacity local data storage;
• Overall OEE statistics and energy analysis;
• Quality analysis and optimization applications;
• Predictive maintenance application scenarios;
• Professional process data analysis tools and applications;
• Interaction with cloud-based application systems;
• Integrated monitoring and business intelligence.
4 Examples of Hypervisor Application Scenarios
(1) Architecture of distributed energy storage system
In a distributed energy storage system, an industrial PC supporting a hypervisor, namely APC910, is used to operate the entire system. This includes tasks such as controlling battery racks, inverters, and I/O via CAN/ModbusTCP, POWERLINK, etc., as well as managing the energy storage system running on Linux, metering, analysis, trend charts, alarms, logs, and other management tasks such as cloud platform connection, building control, and remote maintenance, as shown in the application architecture in Figure 4.
Meanwhile, the system needs to store 40MB+ of data daily, which is then used for learning and optimization of data analysis software on the Linux platform.
(2) High-end and sophisticated HMI design for injection molding machines
Windows/Linux is incredibly powerful at handling display tasks, and it boasts a vast array of controls developed by IT professionals. Using these controls in machine display and operation can truly elevate a machine's "high-end" and "sophisticated" appearance. Therefore, B&R's mapp VIEW has been extremely popular with users. It supports animation, video, PDF files, and various IT control display technologies.
Figure 4. Distributed energy storage system based on Hypervisor architecture
Figure 5. Application scenarios of mappView on injection molding machines
In Figure 5, the Smartmold 9 controller, which utilizes Hypervisor technology, displays a beautifully designed and smooth interface developed using mappView on Windows. Simultaneously, real-time tasks such as mappPLASTIC, mappHydraulic, and mappAxis also run on the Smartmold 9 controller, achieving a seamless integration of machine control and the Windows HMI.
In fact, besides using Hypervisor technology in HMI processing like mappVIEW, the optimal parameters for the entire injection molding machine can be obtained entirely based on this architecture.
