Partnership Combines Cybersecurity With Predictive Maintenance

Partnership Combines Cybersecurity With Predictive Maintenance

Data Analysis: At the Edge or in the Cloud?
Industrial Ethernet
From CPU to the Cloud

Bringing data analytics and artificial intelligence together for IT/OT insights, Atos and Nozomi Networks provide value whether deviations are caused by malicious or non-malicious events.

Much of the news coming down the pipe these days about
cybersecurity—other than the news about the cyber attacks themselves—is
of a collaborative nature. More companies throughout the cybersecurity
ecosystem are finding a need to join forces to help industry fight off
the bad actors.

Although on the surface a recent announcement about Atos and Nozomi Networks teaming
up looks to be more of the same—a classic cooperation between an IT
heavy hitter and a cybersecurity company focused specifically on the
industrial control space—the partnership highlights an even more
important development. These two players are bringing together data
analysis capabilities not only for real-time cybersecurity protection
but to provide the predictive maintenance and operational analysis that
the increased visibility makes possible as well.

This is of particular importance for manufacturing personnel
struggling to make an argument for the budgets needed to protect the
operations. Given a choice between investing in what is essentially an
insurance policy vs. investing in innovations that can move the business
forward, executives tend to lean in the latter direction.

“Cybersecurity investments are always difficult because they don’t
bring in top-line revenue,” noted Mazi Fayazfar, chief technology
officer for telecom, media and utilities at Atos North America. “When
you’re going to your board of directors, and you’re trying to justify
these investments, the way you justify those investments have to be
fear-based, they have to be risk-based. They have to determine how much
of the total money that I have to spend on moving this business forward
do I want to invest in an insurance policy? Or how much of that dollar,
that share of wallet, do I get to spend moving my business
forward—innovating and driving new experiences for my customer?”

As a result, too many manufacturers are impacted by not spending the
money necessary to mitigate the risks. “We really see there’s a value
here in being able to invest in a toolset that gives you the ability to
sleep at night, but isn’t singularly a liability, but it is also an
asset for you,” Fayazfar said. “And being able to turn that investment
into dollars that move your business forward while also helping to
secure infrastructure.”

Atos Codex is Atos’ brand umbrella for everything in its analytics
portfolio—from consulting on outcome-based projects to a
high-performance computing platform. The company is able to analyze
structured and unstructured data from different sources, such as
operational networks, and provides robust insight to users across the

Nozomi’s key strengths in this partnership are its knowledge of the
manufacturing space and its focus on cybersecurity for industrial
control systems (ICSs). Its fourth-generation SCADAguardian uses
artificial intelligence (AI) and machine learning to provide real-time
ICS monitoring, industrial network visualization, hybrid ICS threat and
anomaly detection, asset inventory and vulnerability assessment.

Although SCADAguardian is focused on cybersecurity concerns, that
data crosses over well to predictive maintenance, Fayazfar noted. “A lot
of the machine learning and the artificial intelligence of what Nozomi
does is focused on building a baseline of what that data looks like,
what values happened when and with what consistency does that continue
to happen,” he said. “And it’s really that base of knowledge that gives
you a baseline in order to determine deviation.”

Sometimes that deviation is caused by malicious intent, but other
times there is no malicious intent involved. “Sometimes that deviation
is a function of degradation of components over time—the pitting of a
contactor, the water entering into hydraulic lines and seeing braking
systems start to slow down, or motors starting to fade or the brushes
starting to fade and we’re starting to see a degrade in performance,”
Fayazfar said. “But the metrics, if you’re measuring to determine those
values coming back, are the same whether you’re looking for a deviation
from a malicious or a non-malicious event.”

The joint solution from Nozomi and Atos will help customers decrease
total cost of ownership while implementing both analytics and
cybersecurity within industrial networks. “By implementing these tools,
you get all of that visibility that you would otherwise be using
strictly to secure you in an insurance context,” Fayazfar said. “Why not
leverage that investment you’re making to get real results today that
can improve your business, extend your assets or give you potential
upside if you know how to monetize it?”

At the Disney Data Analytics Conference next
week in Orlando, Fla., Atos and Nozomi will make a joint presentation,
“Deep Neural Networks: Applications in Operational Technology.” Tim
Oates, principal data scientist for Atos, will educate the audience
about different types of AI, reinforced learning, machine learning,
supervised and unsupervised learning techniques, and neural networks.
The second half of the presentation will have Nozomi co-founders Andrea
Carcano and Moreno Carullo explaining the practical application of those
AI techniques in an ICS environment. The two companies will also demo
their joint solutions for the first time publicly at the Atos booth #208
during the conference.