ILMA Compoundings June 2018 - 25

or many of us, the power of artificial intelligence (AI)
is as close as the nearest smartphone. With simple voice
commands, we can enlist Siri, Alexa or Google Assistant
to dictate and send a text, tell us the weather and even help us
make yes/no decisions by flipping a virtual coin.
These are still early days for consumer-facing AI, though.
Digital assistants are as likely to frustrate as they are to
"assist." And not every smartphone owner is an AI user;
many still don't understand what value AI-powered digital
assistants can deliver.
The situation is similar for lubricant manufacturers
looking for digital assistance in running their businesses.
Some see great potential ahead for AI-driven applications to
recognize and act on speech commands. Others believe AI
can help them generate and analyze data with a speed and
comprehensiveness that would be highly ineffective if tried
by humans, if not impossible. Still others believe AI will
enable them to become better-informed decision-makers
through more accurate forecasting. Despite this, currently
there are few industry-specific, AI-driven applications on
the market. And few lubricant manufacturers are actively
seeking AI-based solutions.
Part of the challenge to a widespread push for better
AI integration at the manufacturing plant level is that
"AI" encompasses a wide range of possible processes and
applications, with varying degrees of hardware and software
integration. Its use in one plant might be different from
how it would best serve another operation. Plus, industry
leaders agree that, like their smartphone-owning counterparts, lubricant manufacturers don't typically go shopping
for AI. Instead, they shop for solutions - solutions to
current challenges, improved efficiency and performance,
and better data in support of better decision-making. What
they might not realize is that AI functionality can address all
of those issues.

What Is AI?

"AI is a very broad topic," noted Annie Jarquin, Londonbased director of business development for Kline Group.
"You have pattern recognition, data analysis, cognitive computing. It's Industry 4.0" - a designation that also includes
cloud computing and the commingling of computer-based
and physical systems.
"It's a big topic that gets confused with so many points
along the supply chain," she said. "It helps to narrow down
what it means to you and your customers. That's a critical
first step for companies."
"Of all my customers, no one has asked for AI," said Sean
O'Donnell, chairman of Datacor. "Not because they don't
want AI; they just don't view it as being AI. They just see
the analytical ability as a feature function of the software."

In addition, when it comes to AI-powered applications,
"customers don't know what's available," acknowledged
Balaji Athimoolam, controls engineering manager for
Emerson's Blending and Transfer Solutions. "They simply
come to us and say, 'We have lots of archived data. How can
we make the most of it?'"

Pinpointing the Potential

Before answering that question, Athimoolam continued,
"We ask them what results they want. Mostly, they're
interested in improving their blending systems and operations. Data analytics can help them make certain they use
their equipment efficiently, achieve the best throughput or
identify what they are missing out on due to bottlenecks."
Data analytics, however, is just one possible use for AI
software. "I think [another] significant role is in pattern
recognition," O'Donnell said, "being able to sift through
accumulated data and use that information to guide future
decisions and make decisions as a human would. Today's
systems are looking at past customer behavior, purchase
orders, batch size, etc., in order to set the parameters for
calculating material requirements planning [MRP] and
production scheduling. Traditionally, data-entry heavy
applications and those that required a lot of human analysis
are now using AI to look at past patterns of behavior to
come up with decisions moving forward."
Athimoolam points out that much human analysis
has been hampered in the past by the way information is
typically segregated in manufacturing plants. AI has the
potential to overcome that hurdle.
"In the lubricants blending industry, there is a large
amount of data available from the existing controls systems
used in lubricant plants. But all the data is available in silos.
There are not tools readily available to analyze that data and
make inferences to help in the decision-making process,"
he said. "AI can be a tool to make better use of the data
available from different historians to provide predictive
intelligence to better utilize plant resources.
"Almost all clients ask us, 'How can we optimally use our
available equipment? What is preventing us from making
the best use of our equipment?' For example, most of these
plants have periodic equipment maintenance schedules that
require them to shut the plant and service the equipment,
even if certain pieces of equipment were not in use enough
to warrant that kind of maintenance. With AI and smart
instrumentation, we can take a precise look at what equipment got used to the point where periodic maintenance
would be beneficial. Plants can better manage and better
plan their maintenance around their production schedule."
Jarquin is convinced that the data will be vital in other
areas as well. "I think AI growth will come more in the

ILMA Compoundings June 2018

Table of Contents for the Digital Edition of ILMA Compoundings June 2018