Air Liquide digitalizes to compete and serve customers

Dec. 3, 2018
New digital tools are helping to elevate worker tasks and redefine organizational structures.

By Jim Montague, Control executive editor

Even if they’re desperately needed, big changes can still be very hard to pull off. This is doubly true for

Air Liquide's Arnold "Marty" Martin

epic projects like implementing digitalization in manufacturing processes because they affect so many people, devices and systems, and because they typically represent a huge shift for all these entities from the ways they typically worked for years and decades.

One company facing these challenges head on is Air Liquide, which is embracing digitalization on several levels to help it compete more effectively, and work more closely and efficiently with its customers. The $25-billion air-separation firm is the global leader in small-molecule gases and related technologies, and it serves users ranging from hospitals and home healthcare patients to semiconductor fabs and oil refineries.

“We need a step change in operations excellence, and we need it to evolve and get to the next level,” said Arnold “Marty” Martin, director, process control technology, Center of Technical Expertise (CTE), Air Liquide. “We’re really a utility to our customers, so we absolutely have to be there for them. When we sneeze, our customers get a cold.”

Martin presented “Air Liquide goes digital to transform customer intimacy” at Smart Industry 2018.

NEOS leads to SIO

To prepare for the step change it needed, Martin reported that Air Liquide recently developed its year-old NEOS (Greek for “new”) program, which calls for the company to achieve €1.2 billion in efficiency gains by 2020. The program will draw its staff, customers and all other stakeholder closer together; support operation in all its different economic environments; and enable further innovations in its core businesses.

“We have to change because our customers and competition are changing, too,” explained Martin. “We’re an old company, and we need to do more to show our customers that we’re there for them and can give them they options they need. We can’t stay No. 1 by sitting still and resting on our laurels. We must keep leading the pack, so we’re investing a lot of money and people on this big transformation, and the numbers coming back so far look very good.”

A key component of NEOS is the company’s Smart Innovative Operations (SIO) initiative for managing its many plants, systems, equipment, valves and other assets, as well as interacting with customers for greater reliability and cost effectiveness. Martin reported that SIO aims to help Air Liquide achieve 99.99% reliability by predicting and dealing with problems before they occur.

“This isn’t about gathering more data,” stressed Martin. “We’ve got tons of data. This is about what we do with that data.”

The company will take its many data sources, and add analytical capabilities. These include software for enterprise information management, business intelligence, enterprise performance management, data warehousing and governance, and risk and compliance. This will be followed by giving users easier access to the results for better collaboration, and carrying out services and best practices.

Remote monitoring key

Martin explained that operational excellence for SIO at Air Liquide includes:

• Connectivity that can leverage expertise

• Proactivity with early detection and case management

• Digitalization with secure access to data

• Sustainability for continuous improvement

“This means a foundational shift, and moving to control systems that operate our plants with a lot more remote monitoring,” said Martin. “Eventually, human intervention in processes will be treated as an incident to be investigated. Over time, we’ll have partially or minimally attended plants, which will also have centralized monitoring of operations with evolving roles and responsibilities. “This will be possible with the IIoT and as other new technologies become available. However, we’ll need a true roadmap—not just what’s the latest programmable logic controller to adopt—so we can make sure we’re ready for these next-generation challenges. We see the five levers that will enable this transformation as: plant automation, data management, data analytics, operations digitalization and asset management.”

People must change, too

Of course, along with all these technical advances on the way to operational excellence, Martin added that Air Liquide’s personnel will have to upgrade their understanding, habits and skills, too. “Operators have got to evolve or they’ll be eaten alive,” said Martin. “IIoT is just four letters if you don’t know what you’re going to do with it or how your plant could use it.” To begin this more difficult transition, Martin reported that Air Liquide has outlined several primary tasks:

• Knowledge capture, because veteran operators won’t be available indefinitely, and virtual operator assistance, advanced process control (APC) and multivariable process control (MVPC) tools will be needed to help guide less-experienced replacements.

• Cultural issues, because complex, new software requires training. Personnel and their needs must be understood because the best technical solutions will come unglued without them.

• Company accountability and agility, because new business environments require keeping pace with technology. IIoT can make data available, but users must learn to turn it into usable information that can make a difference in operations. “We have to train people before we can begin to hold them to a higher standard,” added Martin.

New, cool, hybrid roles

As remote monitoring and minimally attended “many mini” plants increase at Air Liquide, Martin added that several interesting roles are going to emerge due to SIO that combine skills from formerly separate jobs. “If we’re going to have minimally staffed plants, it’ll be even more important to cover the tasks that remain,” explained Martin. “We’ll need production people that can not just look at data, but analyze trends more thoroughly and understand where their plant is going. We’ll have to pair experts with less-experienced staff, and teach them to understand and make better decisions. And, we’ll also need multi-skilled technicians able to work on more different types of equipment in more applications.” Martin added these new roles and opportunities will likely include:

• Real-time pilots (RTPs), who will perform remote monitoring of operations through service-level agreements (SLAs). They’ll monitor and respond to alarms, and collaborate with on-call personnel as needed. They’ll also handle equipment lists and process variables authorized for RTP responses.

• Operations analysts, who will perform data analytics and case management. They’ll use data analytics to monitor for areas of improvement using smart key performance indicators (KPI); case manage issues detected; optimize remotely performing predictive analytics on key equipment; and help prepare production and energy supply for customer demands.

• Multi-technicians, who will be more analytical and multi-skilled, using data to optimize decisions, conduct remote monitoring and control tasks, and perform some maintenance activities.

New organizational structure

Beyond reshaping individual jobs, Martin added that SIO will also drive a new functional structure for his entire organization. For example, product supply manager will deliver production volume setpoints to the RTPs, who will provide case management information to a centralized team of analysts. The company is also implementing new virtual operator assistant functions by standardizing on Platform for Advanced Control and Estimation (PACE) software from Yokogawa Electric Co. “I put in MVPC 20 years ago, and done right, they can be great,” added Martin. “But once MVPC is in place, the operators may not know as much as before about how the plant works because the MVPC is taking care of it.” “We can make SIO work, but we need the solid tools we’re developing. We also know this process can be frustrating for our plants, which is why we’re also developing the plans and resources to make it happen, and spending wisely about where we want take our plants. We do a lot of replacements because many plant were put in 20 years ago, and they need modernized controls, but they also need enterprise historians, cloud services, virtual backup and artificial intelligence (AI), which we’re also beginning to explore.”