Mobile Navigation

Automation & Control

View Comments

Workforce 4.0: The Human Side of Digital Transformation

| By Scott Jenkins, Chemical Engineering magazine

As chemical process industries (CPI) companies continue to experiment with, invest in, and implement a host of digitalization tools, workforce engagement and involvement is the key determinant of success

Chemical process industries (CPI) companies are entering a critical stage in the movement toward digitalization (Industry 4.0), in which the majority of organizations are now initiating pilot projects aimed at improving operations with advanced digital tools. This includes a wide range of technologies, including data analytics, cloud computing, machine learning, artificial intelligence and many others. As the digitalization transformation of the CPI gains momentum, it has become clear that the movement is as much about people as it is about technology. The acceptance and involvement of workers is critical to the successful adoption and expansion of digital tools, as they are asked to adapt to new work practices.

Greg Smith, a senior consultant at the Cutter Consortium (Arlington, Mass.; www.cutter.com) who has worked with companies from across many industries on digital transformation initiatives, says “What I’ve found is that you can always make the technology work, but the ultimate success or failure of a digital initiative is always tied to the people.” He emphasizes: “Companies don’t adopt new technologies; people do.”

Culture of adoption

The CPI has entered a critical stage in its evolution, where most companies are experimenting with the implementation of digital technologies, and are increasingly running pilot projects to determine how best to utilize them. In doing so, organizations are grappling with the reception of those new technologies by the workers who will have to use them, and are finding ways to engineer how the new technologies are incorporated by existing personnel. Smith talks about creating the right culture for adoption, an approach his company calls “adoption engineering.”

“We tend to assume that humans will behave rationally, but that is only sometimes true,” Smith says. “Workers’ previous experiences, along with a large set of preconceptions, existing biases, heuristics, shortcuts, and so on, all color their view of a new technology and affect how they will receive it,” Smith says. The objective for companies then becomes how to create conditions that will motivate employees to embrace the technology.

Resistance to the introduction of new digital technologies can come from different sources. Some workers are threatened by change itself, and are not inclined to do something new, even if benefits are possible. Others resist externally imposed changes.

Additionally and importantly, resistance comes from a conception persistent among many workers that digital tools will make them less valuable to their organization. “The biggest hurdle we see is that workers have a lot of fear that digital technologies will result in job loss,” says Jane Arnold, global lead for process control technology at Covestro AG (Leverkusen, Germany; www.covestro.com).

But Arnold says the reality is different, and she’s not alone. She and others view digitalization as an opportunity to liberate workers to undertake more valuable, higher-level functions that involve more creativity and imagination.

“Digitalization offers a way to change the way we work; the intent is not to reduce headcount,” she says. “We have to maintain the human element in CPI operations. There are important differences between chemical processing and repetitive discrete manufacturing. The CPI needs many engineering, maintenance, operations and other functions that will not be eliminated by digital tools.”

Michael Risse, vice president for advanced analytics software provider Seeq Corp. (Seattle, Wash.; www.seeq.com) also sees concern on the part of some workers that digital technologies will replace job functions, but stresses that the fears may be overblown. “Digital technologies will enable workers to do things that haven’t been possible before — that’s the defining experience with Seeq. Applications reduce the time required for certain tasks so workers can devote time and intellectual resources to areas that could be extremely valuable to the company, but that in the past, were not addressed because people were devoting their attention to routine operations.”

The idea that digital tools are a liberating force for workers in the CPI is also shared by Rajiv Anand, CEO of Quartic (Oakville, Ont.; www.quartic.ai), a provider of artificial intelligence solutions for the minerals, chemicals, petrochemicals, pharmaceuticals and food and beverage industries. “Artificial intelligence (AI) and machine learning allow engineers to move toward more proactive analytics, rather than simply investigative, root-cause analysis work,” he says, adding that this results in workers having a more visible impact of their work sooner.

AI with machine learning tools also allow engineers to “zoom out,” Anand says, by which he means looking at the interactions within a process at the level of a unit operation or at the full plant, rather than at an asset or loop level. “This is particularly useful for processes that have a lot of inherent lag and are highly interactive,” Anand notes.

Taking the idea of digital’s freeing quality further, Covestro’s Arnold views digitalization as a continuous improvement exercise. “If you are able to solve one problem by using digital tools, you free up personnel to work on the next problem that would have just been neglected, or tolerated or ignored. “People add value; so we need to utilize them,” she says.

Duane Dickson, vice chairman and U.S. Oil, Gas & Chemicals leader, Deloitte LLP (New York, N.Y.; www.deloitte.com) says an ideal model would be to have digital tools deployed to help people do their jobs more safely, faster and more efficiently, then utilize the surplus person-hours for solving commercial and technical problems that companies don’t normally have time to address because they are devoting all resources to producing products. “This scenario does happen,” he says.

Employee engagement

In addition to overcoming fears, the adoption and overall success of digital tools also depends on the engagement and involvement of employees with the development of the digital solutions themselves. “Nothing will be adopted by end users unless they are involved in developing the solutions,” says Covestro’s Arnold.

Covestro recently began a pilot project at a site in China (Caojing Chemical Works) with applied artificial intelligence (AI), where the company uses machine learning software to recognize patterns and deviations and to predict equipment failure up to eight months in advance, and avoid unscheduled downtime. “The project is about AI, but just as much, it’s about how to get employee engagement right, because without employee engagement, no digital technologies will work,” Arnold says.

The pilot project is beginning with 20 volunteers representing all different functions within the plant, including maintenance, operations, plant manager, and so on. The team is working on revising work practices and building a library of information that employees can use to quickly assess the health of an asset. “We need to figure out what makes a system intuitive for workers. What do they like? What would make their work easier, faster or better? What are their pain points?” Arnold says, adding “We want to help them, not just give them another screen to look at.”

This approach allows each site, plant, unit and team to customize the digital solution to allow for nuances that differ from the others — different ‘personalities,’” Arnold says.

Cutter’s Smith says the way to undertake smart digital transitions is to give workers freedom to try different things and quit the things that don’t work. A corporate culture where “failure is not an option,” does not help in digital transformations; it only hinders.

Smith remarks that engineering-driven companies and science-based industries sometimes have a hard time adopting digital technologies because it often requires them to give up some hierarchical control. “Companies need to start small, generate some validated data, then make incremental changes. They need to involve their workers in those incremental changes, and in doing that, will have to embrace a messier distribution of authority than they are generally used to,” he says.

Another aspect of engaging employees in the development of digital solutions comes from the deployment of machine learning algorithms. Quartic is looking to overcome that lack of trust that occurs when AI models are built by one group and used by another. “Most of the digital transformation in this [AI] context is led by IT [information technology] and data science people,” Quartic’s Anand says, which is a large barrier. When this happens, machine learning algorithms can appear to be a “’magical black box’ that is doing something we don’t understand, and therefore, we don’t know if we should trust the predictions,” Anand says.

The best way to overcome this is to let those who will be impacted by the technology build the AI models. The tools have to be built for that workforce so they can develop their own AI applications, argues Anand. Further, employees should not be forced to learn new tools or techniques — instead the new tools should become a natural part of their workflow.”

Quartic provides a platform that enables process manufacturing subject matter experts to build intelligence with their own knowledge, and with tools that look and feel like the operational technology (OT) systems that they are already familiar with, Anand says.

Experiential learning

In 1983, the British psychologist and scholar Lisanne Bainbridge outlined the “ironies of automation,” in a paper that argued that while highly automated systems may require less human input overall, the interaction with the system that humans do have becomes more important. Although the idea predates the use of advanced digital tools and AI industrially, it still has traction in some contexts.

Don Glaser, CEO of Simulation Solutions Inc. (Shrewsbury, N.J.; www.simulation-solutions.com) mentions Bainbridge’s work when he remarks that among the challenges associated with digital transformation are that, as the level of automation increases, the actions required of humans become more consequential, and the skills and practice of workers becomes more important.

When process operators and engineers are younger and less experienced, they are less well equipped to respond appropriately when a process upset occurs and the automation tools break down, Glaser says. At some point, people will have to manually intervene to break the automation loop, but if they haven’t practiced, they don’t have confidence in manipulating the process, he adds.

Becoming more reliant on digital tools can sap humans of process knowledge, but digital tools can also help solve this issue. Duane Dickson at Deloitte says that part of the solution is to increase hands-on learning. Companies need to make digital technologies more accessible, so that employees have opportunities to play around with the tools offline, and get familiar with how they work and how they interact with the process. “The more familiar they are, the more likely they will be to give buy-in for the implementation,” he says.

In an example of experiential training, Don Glaser’s company Simulation Solutions offers “active learning environments” designed to provide simulator-based training that fills gaps in hands-on learning. “Chances for experiential learning are lacking at all levels — operators, engineers, supervisors,” Glaser says. “For young engineers, there is an operational sense that is insufficient, and the biggest gap is in experiential learning.” Workers need to understand how processes work, so they can intervene when the automation fails.

“Simulations are designed to ‘de-instrument’ the process, so workers get a better idea of what the automation is designed to do,” Glaser says.

Another example related to hands-on training appeared in September, when Emerson (St. Louis, Mo.; www.emerson.com) introduced its Performance Learning Platform, a portable and compact automation-technology training solution that enables hands-on training to prepare workers to maintain plants safely and efficiently. The platform reinforces competencies essential to fostering digital transformation and helps close the workforce skills gap, Emerson says.

Industry-academic partnership

Industry-academic partnerships are proliferating to fill in some of the gaps in experiential learning as well, both to train students in using digital tools, but also in using digital tools to train students on conventional process equipment. An example of this can be found at a center opened this autumn at San Jacinto College (SJC; Pasadena, Tex.; www.sjcd.edu), a school located in the Houston ship channel, home to dozens of CPI plants.

In 2015, the East Harris County Manufacturer’s Association developed an initiative for realizing a vision of an enhanced workforce for process plants called PetroChem Works. Among the outcomes of the partnership was The Center for Petrochemical Energy and Technology (CPET) at SJC, which was opened in August 2019, with 29 custom laboratories (six more are still being built). CPET is designed to foster hands-on learning, as well as for exploring ways to leverage digital tools to change curriculum, content, evaluation and more, says SJC associate vice chancellor Jim Griffin.

CPET has enjoyed close collaboration with equipment companies and operator companies since its inception, Griffin notes, citing an example in which operations personnel and trainers from LyondellBasell (Rotterdam, the Netherlands; www.lyondellbasell.com) came to the school to work with students on distributed control systems, and in how to look at data and make decisions about dealing with different process scenarios. Griffin says representatives from other CPI companies have worked with students on mobile digital field devices, wireless transmitters, making decisions with trending data, digital alarm management and other topics, and Emerson Automation Solutions has provided automation equipment to the center’s fully functioning glycol-processing unit.

“We are continually looking to add capabilities to use AI to improve teaching and learning,” Griffin says.

Gamification of learning

In another industry-university collaboration designed to provide realistic training, Columbia University (New York, N.Y.; www.columbia.edu) professor Robert G. Bozic is working with Simulation Solutions chief engineer Matt Garvey and CEO Don Glaser on a simulated process safety exercise that has been “gamified,” a concept that refers to the application of elements from video gaming into work, training and education tasks.

Bozic and Simulation Solutions have developed a competition called ChemE-Sports, in which chemical engineering students compete with peers to earn scores on a computer simulated process safety and optimization unit. As part of a four-day chemical engineering operations exercise, the students operate a simulation of a distillation unit operation, and are required to act as control room engineers and operators adjusting, controlling, observing and taking action when upsets, alarms, and potential safety situations are introduced. Teams earn points maintaining the profitability of the process and by keeping the number of alarms and time in alarm low. Inspired by the rise of E Sports (competition using video games).

“The students have an on-screen representation of the equipment that mirrors what they would see in an industrial control room,” Bozic says, but without real risk.

Bozic, Glaser and Garvey led a competition recently at the AIChE fall meeting in Orlando, where teams of students from different universities competed against each other in a knockout tournament format. “After the in-class competition, we take time for after-action review,” Bozic says, “which forces students to reflect on how the process engineer responded to issues and what actions they took.”

Other examples of gamification include scenario plays, in which a machine-learning model is built and then presented with data for different scenarios and the outcomes observed, Quartic’s Anand says. “For engineers and technical people in the CPI and other process industries, gamification has to be done in context. Scenario plays is one of the most useful ones that we have experienced.”

Competition also works. “We do reverse hackathons — where an AI model is built by an expert, but you then challenge the peers to improve it or “crash it” based on their knowledge of the subject,” Anand says.

Cutter’s Smith told of another example where a water utilities organization created a “treasure hunt” for finding unknown missing legacy corporate assets (old underground piping systems) using digital tools.

AI mindset versus engineering mindset

In order to get the most from AI, some “de-learning,” rather than learning might also be in order. Quartic’s Anand explains: “As engineers, we are all programmed to think logically, come up with a logical answer to a question and then ‘program’ it into some kind of a logical system,” he says. “AI works quite the opposite way — so you have to be open-minded and focus more on asking the right questions and letting the data find you the answers.”

He continues, “The people who will be successful in AI in the CPI will be the ones who can ask the right questions — if you can ask your business question in a way can be answered by AI, you will have a good chance of making it work for you. Try to be an imaginer, rather than an engineer.

“Don’t try to come up with the answer or you will end up overfitting the algorithmic solution, you will add your bias to it,” Anand says.

Questions — how good you are at asking very specific questions is what will make you successful. When using AI successfully, can you define a problem well? matters more than can you design a solution?

How do you translate pump leakage into an AI-addressable question? Turning a big problem into a specific question that will inform the model that can be answered by AI – that will be the path to success.

An algorithm with reasonable accuracy that addresses a specific question is better than an algorithm that is 100% accurate, but doesn’t get at a relevant question.

Personalized portal

The digital transformation in industry is being driven partially by the fact that digital technologies are increasingly found elsewhere in people’s lives, so the level of familiarity is higher now than ever before.

Emerson’s new personalized digital experience, called MyEmerson, accelerates digital transformation by connecting people and technology through streamlined work processes and better collaboration. With a MyEmerson online account, users can access digital tools to quickly engineer solutions, manage software and installed assets, access training, collaborate with experts, streamline procurement processes and improve visibility into buying history and trends.

“Driven by our personal interaction with digital technology, customers have new expectations today about speed and access to information,” said Brad Budde, vice president of digital customer experience, Emerson Automation Solutions. “Our customers still want access to human expertise, but now expect a great digital experience as well. Combining these two experiences to deliver information immediately and use it to solve problems faster is what drives new business value.”

Digital engineering tools help engineers collaborate, gain confidence in an evolving industry, and streamline time-consuming manual processes. With online sizing, selection and configuration tools for measurement instrumentation, valves, actuators, fluid control, pneumatic and electrical solutions, engineers can confidently and accurately specify solutions for their unique requirements and process conditions. By employing digital tools, engineers can configure instrumentation up to 93% faster, typically saving over 100 engineering hours annually.

Scott Jenkins