Digital transformation in a multi-facility collaboration

Manufacturers with complex facilities scattered across multiple locations quickly learn two facts about helping people effectively work together:

Connectivity is not a function of distance. Two process units can be at a single location and still completely lack interconnection and interaction.
Collaboration, when done effectively using digital tools, does not depend on proximity. It can span any distance.

Provided people are willing to engage, what are the best available digital tools, and how do they make a difference? A digitally transformed enterprise will typically deploy a collaborative-information (CI) server designed to support integration of all facilities, systems and equipment across scattered manufacturing sites.

The CI server simplifies sharing of data and information, taking advantage of the latest IT technology to extend control capabilities while improving operational and production efficiency. It can be virtualized and deployed on-premise, in a private cloud and in a public cloud. The overarching concept is that digital transformation enables a transition from individual improvement efforts to overall company-wide optimization, explained Yokogawa Electric's Makoto Terami and Chigusa Akana, who explained the concepts as follows:

Emergence of data

As a manufacturing company grows, it typically evolves to a point where complex operations are spread globally across multiple locations. It is common for each site to include multiple production units, which create a mix of intermediate and finished products and transport them from one unit to the next.

The production units were almost certainly not built all at once according to a single master plan. Instead, individual units were likely planned and brought online at different times or through acquisitions, all of which reflect varying engineering practices and technologies. Some are modern, while others depend on legacy technologies. Management and operation of each must reflect its capabilities and condition, which means that optimization efforts, when they do happen, are all too often one-off, individualized projects.

However, this scenario is changing. Driven by the need for increased profitability, plants are required to become more flexible and respond more quickly to changes. Rather than simply executing one-off efforts, companies must monitor and manage integrated operations efficiently across multiple units and locations. This change is supported by digitalization of the plants to create and capture enormous amounts of data, but this data must be organized in a manner that makes it accessible and meaningful.

A typical plant or production unit produces three types of data:

data from distributed control systems (DCSs), programmable logic controllers (PLCs), safety instrumented systems (SISs) and, most recently, from industrial edge devices and Internet of Things (IoT) sensors
operational data, indicating feedstock and finished product inventories, energy consumption, personnel information, from the business side of plant operations that is more often associated with IT systems
asset management information such as equipment run times and health status, which has traditionally originated from control systems but, today, is increasingly sourced by IoT sensor technology.

In addition to real-time information, some of the above-listed systems maintain historical information such as average and totalized process variables over timeframes such as hourly, daily and monthly. Otherwise, given the real-time data, a centralized historian—or the CI server—supports this functionality.

In some cases, digitalization does not create new data but simply automates the collection and consolidation of existing data. This relieves the task of manual collection, while providing greater depth and detail. However, most often, digitalization enables companies to significantly expand the scope of data they can process. The objective is providing the right information to the right people in real time to support effective decision-making. This enables quick response to market changes, cost optimization, quality improvement and improved operational efficiency.

More sophisticated companies use digital twins to optimize production. A digital twin is a virtual digital copy of a device, system, human or process that accurately mimics actual performance in real time. The digital twin is an advanced decision support tool that enables improved safety, reliability and profitability in design or operations through forecasting (what’s next?), prediction (what if?) and optimization (what’s best?). It is executable, configurable and able to respond to changes in the same way as an actual plant, thus allowing various optimization strategies to be tested before implementation.

CI server functions

While the discussion thus far probably sounds very clear, actual implementation is more of a challenge. For many companies, the required networks and tools all exist, at least to a certain extent, but they are likely local and isolated by site and even unit. This brings us back to the idea that “remoteness” is not a function of distance; it can even apply if two units are side by side.

Realizing the most from data analysis and control requires centralization that allows engineers and managers at headquarters to view any data from any location and unit as needed and take appropriate action. This is the primary function of a CI server.

A CI server is a cross-platform control and information system that integrates all aspects of operations, control, data collection, data analysis and artificial intelligence (AI).

While the CI server could be deployed as an on-premise server or high-availability computer (HAC) system in a central location, functionality could be distributed across multiple platforms including the cloud. It could include AI, data lake, historian and other functions or seamlessly integrate with other platforms that provide such functionality.

Providing all this calls for the latest technologies combined with deep experience in control and plant technology. A company wishing to improve its ability to respond to everything from large-scale industry changes to changes to individual processes must keep the big picture in mind.

For example, in order to improve operational efficiency and productivity across all sites and units, it is essential to have consistent collaboration among the operators stationed at each site, as well as with headquarters personnel and specialists in the field.

A CI server links all sites, whether nearby or remote, with all the relevant people in any location. It provides a real-time grasp and overview of production, people, output, inventory and activity at any production site and contributes to improved efficiency, productivity and stable operations at all sites and units.

A CI server supports a variety of external communication drivers and provides an industry-standard open interface that enables the collection of all data, plus interaction with digital twins, where available. This contributes to overall production optimization efforts by utilizing data collected over long time periods and linking it with applications for process analysis and diagnosis, as well as with information and quality improvement systems.

Support for collaborative operations

The problem with an isolated site or production unit is its inability to draw on resources outside of its immediate reach. Improvements of any kind will be limited to the skills and imagination of those on site, or perhaps local consultants. The degree of skill and specialization will, therefore, be uneven and will lead to disparities across a company’s facilities.

To distribute resources more uniformly, a CI server facilitates plant operations through:

locally and geographically distributed plant control from a central location
safe and secure data consolidation and distribution
centralized AI capabilities with access to the large pools of data needed to provide sophisticated analysis.

This results in improved decision-making at all levels, from an individual control-room operator to production and executive management.

As a case in point, access to all remote operations allows specialists such as mechanical and electrical engineers, IT staff and experts to serve the entire company, rather than being limited to a specific site.

From a central location, they gain a panoramic view of the entire company’s operations and can determine where their skills will be most effective.
It’s also necessary to remotely monitor the operations at each site in order to solve problems that are intensified by the increasing scarcity of experienced operators and the corresponding issues of retaining skills and organizational knowledge.

A CI server can connect people wherever they are and enable the provision of remote assistance. This improves productivity for the entire company, while ensuring stable and efficient operations, even at remote locations.

Remote control room

As a cross-platform control and information system using the latest technology, a CI server, in addition to being a data-gathering platform, is also a human-machine interface (HMI) that is able to duplicate everything operators see in their individual control rooms.

Using HTML5, the HMIs allow monitoring and execution of authorized operations from a remote location via any device capable of hosting a Web browser. Support of a desktop or laptop PC, an operator station, a tablet or a smartphone makes it suitable for remote operations.

For example, subject matter experts and production managers stationed at a headquarters can support operators in local control rooms by confirming the process data that they view.

Centralized monitoring is especially useful when specialized skills and knowledge are required for on-site work. It saves time and costs by allowing functions to be carried out smoothly via the Web without the need to dispatch personnel from headquarters. A CI server’s Web-browser environment can also be secured to protect against unauthorized access, data theft, leaks and other cyber threats.

In addition to supporting remote operators, a CI server supports remote engineering from headquarters, or anywhere an engineer happens to be stationed. It also supports global team collaboration, even if the members never see each other, since they can all access the same project information and engineering databases.

CI server architecture

A CI server supports several important communication platforms and protocols common to manufacturing, including:

OPC UA
open database connectivity (ODBC), industry-standard communication
message queue telemetry transport (MQTT), a communication standard that is best suited to the IoT, where a wide variety of data and information is exchanged between devices
IEC 61850 communication standard, ensuring a CI server can communicate directly with a variety of intelligent electronic devices to facilitate the connectivity of manufacturing sites and comprehensive operation systems.

These and other technologies make it possible to connect the wide range of equipment and systems that support the facilities across an entire company. They enable system integration within single sites or units, as well as integrated management across more complex sites. Data from all sources can be stored for extensive time periods on reliable platforms.

AI and data analysis

Creating massive data lakes is not a goal in itself, as the objective is to analyze data and extract useful information to improve operations. Production information, quality information and equipment utilization rates are all related and indicate the current state of a given unit or larger site, while pointing out areas for improvement.

For example, quality improvement software sifts through data, looking for relationships and other indicators that help to reduce failures and improve overall operations. The analysis identifies key performance indicators (KPIs) that can inform operators how well the process is performing and in ways they had not considered previously. New practices can be tested using a digital twin to measure effectiveness and look for unintended side effects of the changes.

There are countless ways in which the data can be analyzed with the assistance of AI software. When creating production plans and KPIs, it is simple to view monthly, weekly and daily historical data, down to hourly or minute-by-minute data if such granularity is needed.

Other Web applications, such as market and weather forecasts and geographic information systems, can identify other factors that could have a temporary impact on production. Is a tropical storm in the Gulf of Mexico likely to affect the output of a plant in Baton Rouge? Will it affect other company sites? How might it influence product and feedstock prices?

In this way, the status of all units at all sites, along with the data affecting each, can be grasped in real time to allow production plans to be revised and executed, as needed, and for whatever reason.

By eliminating data silos and taking a panoramic view of every manufacturing asset, company-wide, it is possible to adjust business-critical production plans and KPIs, while ensuring production and optimization are carried out using flexible plans, and thus be adjusted and revised as needed.

A CI server is not necessarily an all-in-one solution with the magical ability to provide all digital transformation capabilities, but, for a company desiring meaningful improvement on multiple fronts, it can be indispensable.