What comprises a digital transformation? You have to look no further than toward many industries that have been affected—the music industry, transportation, lodging and more.
Emerson’s Jonas Berge described how this digital transformation with the Industrial Internet of Things (IIoT) is affecting the process industries. Jonas specifically looked at the why, how and what of a successful IIoT project.
IIoT projects can lack enough sensors, analytics, wrong technologies, security, too big of a scope, etc. When Jonas works with process manufacturers and producers, when asked about possible architectures, he recommends they build on the standards they may already have in place, such as WirelessHART, Foundation fieldbus and other digital protocols.
When asked about how many sensors required, he counsels that their user requirements should drive this. If the business objectives involve improved reliability & maintenance, the sensors would focus on the plant assets which should be monitored. If the business objectives revolve around energy and emissions improvements, the sensors would monitor energy consumers and combustion zones.
While the process is fully instrumented for control and safety, it is often not equipped with sensors for these business-level requirements. The sensors for these can bypass the control systems and go straight to the historians collecting data about plant operations.
Jonas noted that analytics is a hot topic and often overhyped. They do have a place in some solutions such as predictive process equipment monitoring. Analytics can be run on vibration, corrosion, wall/vessel thickness, temperature and differential pressure. Equipment being monitored might include pumps, compressors, cooling towers, etc.
Platform middleware is typically the historian. It receives data from the data sources and makes it available for the specific software applications that perform analysis and make recommendations. For example, an OSIsoft PI system connects with many data sources and software applications. For more audience-specific solutions, such as for reliability engineers, maintenance managers or energy managers, built-for-purpose web apps connected directly to wireless gateways can provide the analytics from the wireless data sources.
Another question is where to locate the experts to analyze, interpret and make recommendations based on the incoming data and generated analytics? Choices are local to the plant, centralized hubs or expertise, or 3rd party supplier contracted experts on the technologies they design, build and support. The choice is based on the availability of these experts and the strategic nature of what is to be monitored. From a security standpoint, a data diode moves data one-way out of the historian to analytical software packages.
Jonas recommends taking a phased approach and separating IT-focused projects from instrumentation and control (I&C) functions. From an I&C perspective, it starts with identifying the sensors and architectures to move data securely from the process to the historian and beyond.
He concluded with reiterating the importance of building on the platform you have in place and relying on existing standards and taking a phased approach.