Industrial Autonomy on the Horizon

Summary

The era of the remote workforce has brought to light a cross-industry need for resilient, futureproof industrial control systems supporting more efficient, sustainable, and safe manufacturing plants. Through the integration of autonomous software and technologies along the production line, and the innovation of process control systems, plant operators can achieve long-term operational benefits.

Ongoing advancements in process automation have enabled users to create steady-state and dynamic models for plant and control design, assess equipment performance and troubleshoot issues, evaluate process design, and resolve operating problems. By using software, including advanced control and alarm management, industrial organizations have optimized industrial processes to improve business results and safety. This progress has also provided greater visibility into process risk—notably, advanced control-defined optimal limits and rigorously executed controls to maintain those limits.

Today, the introduction of new digital technologies at the plant and enterprise levels has the potential to augment people and processes to an unprecedented degree. Advanced functions, such as artificial intelligence and machine learning, are changing how people work in industry.

Autonomous control systems fulfill the growing need to streamline plant communications, provide support for the next generation of industrial workers, and gain a more comprehensive view of process inefficiencies. Autonomous solutions enhance workforce safety and performance while reducing environmental impact and operating costs with more adaptable and accessible system assets.

Process industries have never been under greater pressure to meet production targets, minimize costs, and maximize asset efficiency, all while ensuring health and safety. Moving forward, companies in these sectors will need to better align production strategies with market demand to maximize revenue growth. They must also embrace the adoption of digital technologies to drive efficiency and make investments in plants to reduce carbon footprints. From safety to sustainability and productivity to reliability, the hurdles presented in the remote-work environment amplified the need for more resilient, interconnected plants.

Yet even as industries grapple with structural changes, and as societies and economies pivot to the “new normal,” process industry companies themselves have a window of opportunity: Now is the time to adapt strategies and technologies to help reduce disruption to operations and achieve new levels of performance and profitability.

Current state

Automation systems in continuous-process plants are constantly evolving due to competitive industry pressures, customer demands, external events, and security requirements. Like it or not, most existing systems have changed as a result of numerous small actions taken over the years. A control system originally installed 25 years ago may include a patchwork of small additions made over time, leading to a system that is difficult to maintain because of all its unique quirks. Only some system owners take a strategic lifecycle approach to their control systems. Others are typically reactive, making changes only as needed to correct problems.

Many industrial sites also suffer from the lack of a consistent philosophy in integrating various plant subsystems. The prevailing information technology (IT) focus on the operational technology (OT) space has only exacerbated this problem.

In addition, the current generation of experienced industrial engineers, operators, and technicians is in the process of retiring. As these workers leave the plant, they take with them valuable tribal knowledge of the control system design and evolution, the production processes, and the associated control strategies. This departure is causing the loss of their collective know-how. Recruiting workers to backfill retirements is just one part of addressing this industrial skills gap. Once new employees are on site, they must be trained efficiently so they can up-skill quickly and produce results.

All of these challenges set the stage for a new approach to the control system of tomorrow. The continuous-process industries are at the beginning of an inflection point regarding what they can do with automation solutions. Today’s objectives should be to leverage decades of process know-how, find ways to integrate subsystems and streamline communications, and become more flexible in how to work with control technology in general.

What is (and is not) industrial autonomy?

The topic of “industrial autonomy” is gaining significant interest, with many diverse views—and compelling opinions—on what constitutes the autonomous operation of an industrial facility. According to a recent study by LNS Research, approximately 50% of industrial transformation leaders have an autonomous plant initiative formalized, and an estimated 41% of these leaders are accelerating their autonomous plant efforts because of the global pandemic. 

The world of industrial autonomy is a crucial part of what ultimately comprises Industry 4.0, which will enable industrial assets and operations with robust adaptive capabilities. Autonomous control systems will respond without operator interaction to situations within a secure, bounded domain that was not preprogrammed or anticipated in the system design.

Industrial autonomy lets industrial companies harness innovative technologies to create a true digital transformation of operational strategies. Because digitization is not a one-step process, understanding a facility’s capabilities, digital maturity, and state of operations is crucial in identifying its next steps in the industrial evolution.

It is apparent that industrial process plants can move on a trajectory toward industrial autonomy and make similar step-change improvements in benefits by harnessing new technology. In an industrial environment, the trend toward autonomous operation is truly focused on optimal advanced sensing and automation technology in plants.

Industrial autonomy is about leveraging technology for better situational awareness. It is about allowing a system to take an optimal action that achieves desired outcomes in the best way possible. And those outcomes are better production, improved quality, more reliable operation and a much more efficient workforce.

Industrial autonomy helps automate a host of plant floor tasks and verify that they are performed flawlessly and consistently. Most importantly, autonomous operations can mean moving humans out of unsafe environments without inhibiting their access or view to process information.

Levels of autonomous operation

When considering the wide range of operational tasks involved in a typical process plant, from the control room to the field to planning and scheduling, it appears that fully autonomous operations may be out of reach for many companies. The process industries will, however, continue to deploy more intelligent, semi-autonomous subsystems that allow the plant workforce to focus on higher-level tasks, even while simultaneously making the operation safer, more reliable and more efficient.

To move toward autonomy in industrial processes, it is important to look at what can be fully automated, what elements will require human supervision, and which areas will remain manual. Once this clarity is established, it is possible to set a path to autonomy following six progressive levels:

• Manual operations: With traditional manual operations, every aspect of the plant enterprise, including instructions and paper-based recordkeeping, is performed manually. Here, no automatic actions occur, with operations relying on humans to make all decisions and perform all functions. Most industrial sites began with significant human intervention required to run and maintain the operations.
• Controlled and optimized operations: Given the widescale adoption of control systems and advanced control software, many industrial process facilities fall into the category of controlled and optimized. But just because most are in this category does not imply that most excel at it. Often an abundance of control loops are still running in manual mode or tuned incorrectly. Control loops running manually or those that are poorly tuned hinder optimization of the process. A large percentage of sites also have advanced control models that do not reflect current process dynamics or equipment performance, leading to poor results. In extreme cases, this situation can result in sites giving up on advanced control altogether. Advanced control and optimization must be viewed as a lifecycle solution—one that is continually kept up to date
• Intelligent operations: The shift to intelligent operations is the essence of Industry 4.0. It is all about software analytics—collecting data, analyzing it, getting recommendations, and taking specific action. Companies can use digital twins to compare current process and equipment performance against expected performance. The benefit of using a model-based approach is that process and equipment performance is evaluated according to known or physical characteristics of plant processes. Users can model changes in the plant and current behavior, see if models are delivering the expected results, and employ the digital twin to close the loop between process dynamics and process design.
• Remote operations: Once process operations are optimized and intelligent, users can think about leveraging the power of remote operations. From remote project execution, service, and support to centrally located remote operations centers, these capabilities are an opportunity to improve workforce efficiency, collaboration, and problem solving, and to effectively serve multiple sites or projects. Remote operations centers are used extensively in the industrial world, particularly in areas with distributed assets.
• Resilient operations: Now, more than ever, plant owners require robust technology to help them withstand faults without having a large number of workers on site. They need resiliency measures such as backup power systems to keep equipment operating. Being resilient means that when failures do occur, the system or operation continues to run normally, and recovery is automated. There are many examples where this method occurs—for instance, in process automation systems with redundancy. Control systems are typically built with redundant controllers. If a controller fails, the redundant partner takes over, ensuring normal operations. At the same time, this controller is now nonredundant, and a process upset will occur if the controller also fails. Resilient operations will reallocate the control application to available system resources, ensuring that operations continue undisturbed, and that system resiliency is maintained, completely automated, with no intervention required by operations. 
• Autonomous operations: Ultimately, the destination of autonomous operations is enabling every day to be the best day of production and all staff to become experts in their assigned roles. The journey to autonomy is characterized by making use of all available digital technologies to realize advances in safety, reliability and efficiency.

For the most part, the far upstream oil and gas business is one of autonomy—wellheads and pipelines are largely unmanned. By reducing the physical presence of people on an offshore platform to the few times it is absolutely necessary, operating companies can dramatically improve human safety while lowering operating costs. The mining industry is also rapidly moving toward autonomy, substantially decreasing the number of people at often very remote process locations in favor of centralized operations.

Although the aforementioned maturity model does have a natural sequence and evolution—it is recognized that most facilities will have a varying degree of capability at each level—plant operators might find that there are manual procedures for some operations, and they also have pieces of equipment that operate largely autonomously. So, it is instructive to assess existing capabilities at each level.

Trending toward the future

The trends are unmistakable: Autonomy is a critical technology that will lead process industry operations into the future. As technology moves beyond automation, autonomy and autonomous systems will bring improvements in many areas.

The latest developments around industrial autonomy provide a timely response to several key industry trends, including the desire for post-COVID-19 preparedness and resilience, growing operational complexity, the aging industrial workforce and upskilling needs.

Regardless of an enterprise’s current tools, design, or talent, integrating autonomous solutions at any level of production serves as a catalyst for increased operational performance by addressing safety, efficiency, and reliability issues to help promote business continuity. An unencumbered vision is necessary to plot the incremental steps to achieve a more autonomous future. This vision requires investing in automation systems in a strategic and consistent manner with the total lifecycle of the plant in mind.

Images courtesy of Honeywell Process Solutions

This feature was originally published in the December 2021 issue of InTech magazine.