Efficient Plant January 2018 - 6
column | editorial
These five steps may help you realize IIoT implementation success.
NDUSTRIAL INTERNET OF Things
(IIoT) is the topic of the day for enterprises that want to move efficiency/reliability
programs toward the predictive and, preferably,
prescriptive levels. Organizations seem to be
organized in three camps:
those that have fully implemented IIoT
technology and practices and are realizing
those that have made an investment and
realized some success, but haven't found a
path that leads to the promised land
those that know they need to get on board
aren't sure where to start and don't have the
resources to experiment.
For the latter two camps, the overriding
question is "How?" I asked Google's experts
for an answer to that question and they came
back to me with a white paper from SAS Institute Inc., Cary, NC (sas.com). The publication,
"5 Steps for Turning Industrial IoT Data into
a Competitive Advantage," lays out one plan
of attack that might help your operations get
in the IIoT game. Download the document at
Here are some high points from the plan:
Step 1. Define IIoT Business Goals: This
pops up time and time again as the starting
point for technology-oriented efforts. Basically, business and technology leaders need to
put their heads together and identify areas in
which IIoT technology can benefit the company. That calls into play the old saw of finding
places where you can realize small victories,
then building on them.
Step 2. Define an Analytics Strategy:
Once you have your use cases in place, select
an analytics platform. According to SAS, assess possible options for how well they deliver
a holistic analytical life cycle that:
efficiently prepares, stores, and transforms data for analytics
drives discovery from diagnostic, predictive, and prescriptive analytic techniques
deploys, manages, and monitors analytics
in the cloud, the fog, and on the edge.
Step 3. Assess the Need for Edge Analytics: This is a key step. According to the whitepaper, IIoT users must do more than analyze
information. They need to turn analyses into
action, which requires a management structure designed to operationalize the insights.
Edge analytics can capture value in real time,
and it deserves special consideration by IIoT
planners. Edge analytics processes the data
stream close to the source of the data. This
allows the analytics system to stem impending problems by shutting down machinery,
triggering alerts, or taking other actions. The
capability for immediate, automated response
is not possible if analysis has to wait until data
reaches back-end storage systems.
Edge analytics also filter data at the source
so that only relevant data is sent to the cloud.
This prevents irrelevant information from
overloading networks and keeps the focus on
what's most important to the business.
Step 4. Choose the Right Analytics Solution: Time to go shopping. This is a big hurdle
and a key decision point, so let the buyer
beware. The white paper offers a good deal of
advice to help you.
Step 5. Focus on Continuous Improvement: As with network security, you can't
take a set-it-and-forget-it approach. According
to the whitepaper, because IIoT continues
to evolve, industrial organizations should
regularly assess their use cases and analytics
performance, and update these areas as new
capabilities and business opportunities arise.
At the same time, they should re-examine
existing deployments to ensure that analytics
continue to achieve use-case goals."
I can't say for sure that this is a one-sizefits-all plan. But, if you're wrestling with
IIoT implementation, the whitepaper could
provide some food for thought. EP