Embedded Technology
The advantages of advanced virtualization technology
By Jarry Chang, general manager of Product Centre, DFI T
he combination of artificial intelligence (AI) and the Internet of Things (IoT) has gradually changed the landscape
of industrial automation. The transformation of the infrastructure of the intelligent IoT system is the starting point of this process. Integrating virtualization, real- time computing, and industrial IoT container technology can also transform industrial automation applications into a “Software-Defined AIoT.” This environment is similar to modern cloud service data centres, reducing operating costs.
The spirit of pursuing smarter production methods to reduce costs and improve product quality has been consistent for the manufacturing industry since the Industrial Revolution. In recent years, the innovative thinking of Industry 4.0 and smart manufacturing has swept the industry. However, AIoT applications cover a wide range of processes and are diverse. A high technical threshold exists for fully integrating the required IT and OT. In addition, most of the solutions of closed machine tool manufacturers are very expensive, making the challenge of profitability and survival even more difficult. The “Workload Consolidation” created by many software and hardware technologies, such as advanced virtualization, has become worthy of consideration. Today, AI has penetrated many aspects of human life, and affecting industrial automation is also inevitable. For example, the ubiquitous variety of colour recognition, size measurement, and appearance defect detection are prevalent applications for “artificial intelligence edge inference.” But this also means that the edge computing device must update the identification model obtained through deep learning at any time. Many deep learning functions deployed on the cloud must also feed back the latest samples from the edge device for continuous optimization and correction. In addition, more and more specific functions
www.cieonline.co.uk.
From fixed function and proprietary to loosely coupled architecture with portable applications & multi-vendor interoperability.
that used to require dedicated hardware have become “software” running on a virtualization platform independently. The result is that the equipment in the industrial automation field will no longer be just a closed system. It is a vehicle that can flexibly deploy various applications in response to changes in demand.
The seemingly mutually exclusive “real-time” and “virtualization” are in harmony
Generally, our most common application of virtualization technology, in addition to the multi-tasking system on a computer, can execute multiple applications at the same time, “travel virtualization” is “system virtualization (or called “host virtualization).” In addition to the “Virtual Machine” for collective isolation, modern cloud service data centres generally adopt the lightweight “Container” technology. Resources are isolated through containers, and only the application program and the corresponding resource environment need to be loaded. Programs do not need to be inserted into the entire operating system, reducing the overhead of computing resources, memory space, and storage capacity.
In the past, “virtualization” often meant an additional performance burden, and it was challenging to convince industrial equipment manufacturers who have long valued real-time
and reliable computing applications to change. However, with the increase in the performance of computer platforms, the development of real-time virtual machine managers, and the advent of time-sensitive networking (TSN) and time-coordinated computing (TCC), it is possible to achieve multi-machine collaboration in a virtualized environment. Accurate synchronization is required, and IT information and OT signals of traditional industrial environments are integrated on the network side, eliminating the need to deploy multiple network environments and thereby saving costs.
Industrial Internet of Things container technology shapes “edge computing data centre” Industries can apply container technology to transfer applications, data, and services from central cloud nodes to edge computing nodes to offload workloads initially in the cloud, such as artificial intelligence and analysis operations. This way, IoT devices can communicate with the cloud in less time, keep data synchronized, respond to system changes rapidly, and even reliably operate independently in long-term offline conditions. Edge devices can also form a “Micro Edge Cloud” through a widespread general-purpose hardware platform, virtualized workloads,
and a centralized container cluster management and scheduling platform. Elastic resource allocation, disaster recovery, and load balancing can realize a cloud-like data centre. If one of the computing nodes is temporarily taken off the shelf for maintenance, the application program initially executed on the node will be automatically transferred to another executable node to continue to operate. If the computing node is offline due to an accident, the work will also be automatically
switched to the available computing nodes to avoid service interruption.
“Software-Defined Networking (SDN)” uses software to change the network architecture and functions to centrally control the entire network and virtualize various types of dedicated network hardware. “Network Function Virtualization (NFV)” is a software application that delivers network functions such as file sharing, directory services, and IP configuration. In that case, it is not difficult to understand the spirit behind the micro-cloud at the edge of the Internet of Things. Industrial equipment manufacturers tend to be conservative by nature. The high price of legacy equipment and the potential liability of having downtime means they are often reluctant to adopt new technology without having a proven use case. However, the manufacturing industry is often faced with “small volumes” and a “variety” of products, so it usually needs to have the ability to adjust quickly and dynamically. The efficiency, productivity, ease of use, and cost-effectiveness of moving to a workload-consolidated system architecture are clear and well-documented, enabling manufacturers to eliminate redundant systems, reduce overall energy consumption, minimize latency, and lower costs.
http://www.dfi.com/ Components in Electronics March 2023 27
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62
