techwatch
BIG DATA ANALYSIS WAS HERALDED AS HAVING THE
ability to offer big insights. It was the tool that was going to
allow logistics and supply chain managers to identify patterns
and connections hidden in piles of data stored in different formats in disparate computer systems throughout the extended
supply chain. It was going to help managers identify new ways
to cut costs, boost revenue, and so forth.
In the world of information technology (IT) jargon, software
that proposes solutions to problems is dubbed “prescriptive
analytics.” Although this might sound like
the answer to a supply chain executive’s
dreams, the reality is that, for now at least,
few if any companies are doing big data
analysis in the supply chain with prescriptive
analytics.
That was the assessment of Michael
Burkett, a vice president for supply chain
research at Gartner Inc. Burkett made his
remarks at the Supply Chain World North
America 2014 conference in a presentation
titled “Big data and supply chain analytics:
Separating fact from fiction.”
Despite the slow uptake of prescriptive
analytics, other types of big data analysis are
starting to make headway in the supply chain.
Most experts classify analytics into one of four categories. The
most common is what’s known as “descriptive analytics,”
insights into what happened in the past and why. Nowadays,
most logistics software applications, such as warehouse management and transportation management systems, come with
some type of descriptive analytics. Because descriptive analytics
has been in use for decades, it’s generally not considered to
be fodder for big data analytics, whose purpose is to explore
uncharted information territory.
Three other types of business analytics are considered per-
fect fits for big data examination. One is the aforementioned
prescriptive analytics. Another is “diagnostic analytics,” which
offers insight into why a procedure went awry. For example,
diagnostic analysis could determine why shipments from a par-
ticular carrier frequently fail to arrive on time at a customer’s
facility. A third category is “predictive
analytics,” in which the software foretells
what will happen in the supply chain. For
example, companies could use informa-
tion on weather forecasts to determine
what consumers might buy, prompting
them to stockpile certain products or ship
them to stores for in-stock availability.
Aircraft and automobile makers could
use sensor-based data
monitoring equipment
to predict when certain
components might fail
and take steps to have
repair parts on hand at
the warehouse or distri-
bution center.
Burkett says com-
panies are already
experimenting with
diagnostic and predic-
tive analytics in logis-
tics and supply chain
applications. So what’s
keeping them from
dabbling in prescriptive analytics? The
reason, according to Burkett, is that the
underlying technology isn’t quite ready
for prime time. It needs further develop-
ment, particularly in the area of artificial
intelligence, which would be required to
help come up with the solutions.
As for what’s ahead for big data analy-
sis, Burkett believes that logistics and sup-
ply chain managers will ultimately be able
to use the results of this type of analysis
to persuade upper management to adopt
new courses of action. “Supply chain pro-
fessionals struggle to bring evidence to the
table,” he said. “Analytics can do this.”
Waiting on big data
“prescriptions”
BY JAMES COOKE, EDITOR AT LARGE
