A Modern EIDI Implementation Approach

Now that the team was ready to go, the first step in the deployment process was to meet with the vendor and the Proclndustries IT team to determine an appropriate architecture strategy that supported usage of the EIDI and its permanent data archive. The architecture also needed to accommodate the many knowledge workers who would consistently access data once they started using the EIDI in their daily routine. Fortunately, the EIDI had a flexible architecture that allowed server load balancing and provided a hardened method of deploying patches and updates, without unduly impacting EIDI server uptime.

IT also needed to anticipate that many more Proclndustries people from various departments would begin querying the data once they learned of its capabilities. They decided to revisit the proposed architecture plan in six months to accommodate future growth. This growth would be likely because a significant number of companies now use more advanced modeling and predictive data analysis technologies, such as artificial intelligence (AI), ML, big data analysis, and business intelligence (BI) tools. All of these new technologies have one common denominator: They require vast amounts of accurate high-fidelity data, which the EIDI would supply.

As the suite of software and control systems used by an enterprise constantly expands, it is necessary to architect such systems in secure environments that take advantage of

  • • The latest and most efficient hardware offerings for processing, memory, and storage;
  • • Fast, reliable networks that are able to transmit large amounts of data needed by people or software systems to analyze information; and
  • • The latest and most effective cybersecurity practices and software tools to prevent unauthorized people or software from accessing EIDI information. Equally as important, no one except for authorized plant operations and IT personnel should have access to plant control networks. Typically, customers place an EIDI on the company business network, where large data queries do not compromise the execution of plant control and safety systems. It also keeps nonessential operating personnel outside of the plant or refinery control network.

Figure 3.3 presents a proven EIDI deployment for a large industrial company, integrating process control systems with an operational management infrastructure. It is composed of

  • • Specialized data connectors and interfaces to specialized control and equipment systems, weather systems, laboratory systems, text files, web interfaces, and geographical information systems;
  • • Real-time data contextualization, visualization, online analysis, and real-time alerting;
  • • Operational data modeling, BI, ML, and process advisory tools; and
  • • A demilitarized zone representing a cybersecurity strategy of separating business and control networks.

FIGURE 3.3

Prov'en EIDI architecture for large enterprises. (Courtesy of OSIsoft.)

The left side of the diagram shows the types of EIDI connectors and interfaces to ingest and store refinery data in control systems: sensors, laboratory systems, along with software or web-based information, and IoT edge devices.

Companies can leverage their existing investments in legacy data collection systems. They are able to integrate all silos of data without having to replace their existing legacy systems (Kennedy 2019). A robust EIDI should be hardware and control system vendor independent. This homogeneous data environment brings new opportunities, such as managing and analyzing operational data from a business perspective (Logue 2019). The existence of a real-time data infrastructure (EIDI) encourages curiosity and innovation, which continuously increases business value.

 
Source
< Prev   CONTENTS   Source   Next >