Virtual manufacturing environment: 4 ways to transition

Here’s what to consider when transitioning to a virtualized manufacturing environment. Look at upgrade timing, assess limits and needs, and minimize failures.

By Chris Di Biase March 14, 2014

Ask IT executives about priorities for the year, and increased use of server virtualization is likely to be on long lists that include security, operating system upgrades, improving collaboration capabilities, and better enabling the mobile workforce. In fact, 53% of IT managers cited virtualization as a top priority for 2013, second only to data protection, in a recent survey of 3,200 IT managers conducted by IT Manager Daily. Virtualization breaks the link between operating systems and the physical hardware on which applications run. Virtualization allows use of multiple instances of an operating system with independent applications on the same hardware, and provides the ability to change hardware without replacing the operating system or applications.

The sharp growth of virtualization technology in recent years is certainly easy to understand, given the cost and efficiency benefits it brings to the organizations that embrace it. Manufacturing companies are about average in terms of adoption of the technology overall-they’ve virtualized approximately 34% of servers, according to a 2011 study conducted by VMWare. Although the technology is steadily becoming more commonplace, the study suggests significant variability in understanding and expertise. These gaps likely indicate operational differences that could be affecting virtualization deployment.

When a popular technology boasting such tangible benefits surfaces, organizations tend to want to quickly hop on the bandwagon. Unfortunately, they often do so without first considering the specific organizational requirements necessary for successful deployment. This is especially critical in the manufacturing environment. Let’s look at four considerations and opportunities unique to manufacturing virtualization:

1. Understanding upgrade cycles

The typical refresh cycle for manufacturing assets is significantly longer than it is for most IT assets. The growth of IT-based assets on the plant floor, however, has made it increasingly difficult for manufacturers to predict and manage upgrade budgets. Traditionally, upgrading IT-based plant hardware such as PC-based HMIs has also required costly software upgrades. Because hardware generally requires more frequent updating than software, manufacturers were often forced to upgrade software prematurely. In addition, each hardware change requires engineering intervention-which can be costly, not to mention the possibility of production downtime it introduces.

Virtualization helps extend the usable life of hardware systems on the plant floor. As a result, it’s much easier for manufacturers to create planned, predictable upgrade cycles. Now, servers can be upgraded without replacing software or involving engineering. This significantly extends the useful life of a manufacturer’s software systems-from 3-5 years to 10-15 years, or longer. What’s more, it allows plant managers to make operating system upgrades when it makes most business sense, rather than because hardware is antiquated. 

2. Knowing your limits

Physical servers housed within a data center are in a generally stable environment designed for a single purpose. Manufacturers face many more limitations when it comes to space, power, cooling, and management of physical servers. Virtualization makes it possible for organizations to consolidate servers, which certainly benefits the IT realm in terms of space and energy savings. The added limitations presented by manufacturing environments actually amplify the benefits of consolidation. To minimize failures among multiple machines, manufacturers generally run one application per physical server. Not only does this cause an abundance of physical servers-all requiring maintenance, generating heat, and consuming energy-but it also results in significant underutilization. The typical x86 server deployment uses only about 10%-15% of total CPU capacity. With virtualization, manufacturers can securely share applications on one server, allowing them to consolidate servers for increased utilization. Consolidation also helps simplify maintenance, reduces the amount of heat generated, and lowers energy consumption costs.

3. Evaluating management needs

The hardware architecture required in manufacturing is more complex than in most industries. The typical shop floor installation often includes a mix of legacy desktops, ruggedized laptops, solid state thin clients, and tablets. Using distributed workstations may offer manufacturers ample computing power, flexible user control, and a fine user experience. But this model also presents a number of challenges in managing workstations, including high operating costs for configuration, patch management, and application support. It also can pose security and data management risks.

Virtualization helps centralize the control and management of operator and engineering workstations, which helps alleviate costs associated with deployment and support of workstation hardware. On the server side, monitoring of all host servers and virtual machines is centralized, and new virtual machines can be deployed using a template. This centralization also simplifies management of installed applications, and users quickly shut down or reset virtual machines without stepping foot on the plant floor.

It’s also possible to extend the benefits of server virtualization to clients. With this approach, software becomes independent from hardware, offering significant flexibility in hardware usage. Instant access to the status of all current clients offers improved visibility across the plant floor. And late nights spent installing software updates and patches become little more than a bad dream-with virtualization they can be centrally deployed. Problem clients can be reset and new systems provisioned in the same manner. Disaster recovery is centralized as well, and security is further enhanced through the ability to centrally entitle or revoke user access.

4. Minimizing failures

Downtime is the enemy in manufacturing, and virtualization is like a white knight. It offers improved availability, which results in increased reliability and lower downtime. With virtualization, machine failure doesn’t automatically equate to downtime as is too often the case with physical servers. Hosts monitor each other and their virtual machines for failure. If a virtual machine fails, it is automatically restarted in a different location where there are available resources. In addition, fault tolerant options make it possible for virtual machines to execute simultaneously on two physical servers so there is bumpless failover in the event of a hardware failure.

Successfully implementing virtualization in a manufacturing environment requires a number of special considerations. Working with trained specialists can help smooth the process.

Those providing automation network and security services can be expertly trained in manufacturing and IT. Such a team specializing in virtualization can provide services and support, including:

  • Assessing the current state of a manufacturing data center
  • Designing and planning a virtual infrastructure
  • Installing, procuring, and configuring a virtual infrastructure.

– Chris Di Biase is principal consultant, network and security services, Rockwell Automation. Edited by Mark T. Hoske, content manager, CFE Media, Control Engineering, mhoske@cfemedia.com.

ONLINE

Rockwell Automation provides more on virtualization services and support.

Key concepts

When transitioning to a virtual manufacturing environment:

  • Upgrade timing
  • Assess limits and needs
  • Minimize failures

Consider this

Assess how your manufacturing virtualization needs differ from IT virtualization.

Learn why manufacturing virtualization can be beneficial below.