It’s easy to get caught up in the exciting promise of the industrial internet of things (IIoT) for industrial automation. The ability to gather and analyze reams of data from across the entire production value chain and use that data to optimize processes—both in real time and predictively—offers transformative potential for industrial enterprises, from manufacturers to energy producers.
But for most enterprises, the old adage holds true: You have to walk before you can run. Before you can begin to tap the potential of next-generation, big data-driven, intelligent automation, you need to modernize the foundation on which it is built. And that means taking a good, hard look at your existing operational technology (OT) infrastructure.
In most industrial automation environments, here’s what you’ll see: musty, old desktop hardware and servers running operating systems and software that are often well beyond their freshness date; dedicated PCs for each application or even for each piece of production equipment; proprietary systems that have been patched and kluged for years just to keep them going.
This legacy OT infrastructure presents a number of problems. First, it's difficult to secure. For systems running on outdated operating systems and applications, security patches may not be available anymore. Even if they are, the “dedicated PC” approach to the infrastructure requires that each system must be updated separately—not just once, but every quarter. This is incredibly time-consuming and expensive, which is why it often doesn’t get done.
The other challenge has to do with connectivity. Data feeds are the lifeblood of the IIoT, requiring lots of connectivity between OT systems and IT resources, including data warehouses, analytics engines, ERP systems, and more. Connecting all the isolated system “islands” that make up the OT infrastructure with the “mainland” IT infrastructure via IP networking creates even more security vulnerabilities.
Finally, there is the challenge of managing this complex, interconnected infrastructure. In addition to the technical challenges, there are organizational issues involved in breaking down the cultural and political barriers between the OT and IT worlds. Who will control what? When something fails, whose fault is it? The inevitable finger pointing takes energy away from what should be the focus: improving business performance.
For all of these reasons, the first step on the path to IIoT really needs to be modernizing the OT infrastructure. You can’t build a solid house on a shaky foundation. Based on my experience working with enterprises across a range of industries, I see a few key priorities to guide your modernization efforts.
1. Modernize, not revolutionize
Some IIoT vendors are urging companies to bypass their OT computing infrastructure altogether. They propose having equipment sensor data communicate with analytics engines in the cloud to perform real-time and predictive analytics to optimize processes and predict failures to avoid unplanned downtime. But for OT teams concerned about even one link to the cloud, now they face the possibility of hundreds or even thousands of such connections. How can they ensure they're all secure, all the time? Moreover, this approach creates a degree of complexity that just isn’t warranted.
Instead, focus on an evolutionary approach to modernization. Replacing those outdated OT systems with new, standards-based hardware, operating systems and applications will go long way toward establishing a stable, secure and reliable automation infrastructure, while minimizing risk.
2. Leverage virtualization
When it comes to just about any computerized environment, modernization means virtualization. Deploying multiple applications on a single, virtualized server residing on low-cost, standards-based hardware reduces the number of physical systems that must be bought, configured and maintained. This dramatically reduces capital expenses and operational costs. It also makes the environment inherently more reliable, because there are fewer physical systems to fail.
3. Build in availability
There is one fly in the virtualization ointment, however. Consolidating applications on a single server also creates the potential for a single point of failure. Of course, this is also the case with the legacy “one PC for every piece of equipment” automation model. But it’s a vulnerability that must be addressed to maintain OT’s always-on requirement.
That means leveraging automation systems that are built from the ground up to provide the desired level of availability required—from high availability and disaster recovery to complete fault tolerance. Making availability a core requirement of your infrastructure rather than an afterthought or an add-on will pay big dividends not only in system uptime but in the ongoing management of those systems.
4. Keep it simple
Finally, adopt a bias for simplicity. Reducing complexity whenever possible helps minimize up-front and ongoing costs, while reducing potential points of failure. Simplifying your architecture using thin-client and other streamlined technologies is a step in the right direction. So is choosing solutions that are designed from the ground up for easy integration, service and maintenance. Many OT teams do not have the specialized training and experience that is common to IT groups in the data center. Deploying solutions that don’t need a lot of care and feeding frees up OT personnel for the higher-value tasks made possible by IIoT, such as using advanced analytics to optimize plant productivity and equipment life in real time.
It’s a journey, not a destination
For most industrial enterprises, the path to the IIoT will be an evolutionary journey. Replacing your outdated automation infrastructure with a more modern, highly available, virtualized environment is the first step on that journey. The beauty of this approach is that modernizing your infrastructure guided by the priorities listed above will deliver tremendous benefits in terms of reliability and manageability right now. Plus, it creates a solid platform on which to build your company’s IIoT future, wherever it may lead.