The Path from Data to Actionable Information as a Driver for the Industrial Ecosystem
These core concepts are crucial to making
the foundational IIoT technologies and
architectures deliver on the promises and
provide the fabric that will enable the new
economies and business models. Never
begin an IIoT project with technology, such
as Artificial Intelligence (AI), as the driver:
Start with a business objective and then
determine the technology and data
requirements that achieve measurable
results and advance one’s strategic goals. To
illustrate the ideas, we will evaluate some
discrete manufacturing use cases showing
how one can deliver rapid business value
integrating open standards from multiple
industrial systems to create rich semantic
context.
O VERVIEW
The foundation of the Industrial Internet of
Things (IIoT) is the notion that abundant
data will be the driver of new products and
economies by providing objective realities
from the production and use of industrial
equipment and systems. There are several
misconceptions about the nature of data
and analytics and how to approach these
problems to achieve actionable and
measurable results. An example is the use of
raw data and deep learning to create
actionable insights. This has led to a great
deal of disillusionment in the industry when
implementing an IIoT strategy.
There is a general lack of understanding
about how to characterize data and
information and what is meant by semantics.
The ultimate goal of any IIoT strategy is to
provide business value. So in this article, we
will look at the requirements from the
perspective of the manufacturing business
and ecosystem to understand the criteria
that lead to actionable information and
interoperability between industrial systems.
We will provide an analysis of relevant
industrial standards such as OPC/UA 1 , STEP 2 ,
EtherCat 3 , SysML 4 , and MTConnect 5 and
differentiate between the syntactic and
semantic standards in greater detail.
1 https://opcfoundation.org/
2 https://www.iso.org/organization/9295.html
3 https://www.ethercat.org/default.htm
4 http://www.omgsysml.org/
5 http://www.mtconnect.org/
T HE P ATH FROM R AW D ATA TO
W ISDOM
The first step is to understand the high-level
taxonomy and the types of data that will be
required to create actionable insights. When
referring to data in the context of IIoT, one is
usually referring to streaming data from
sensors or control systems that are
representing some characteristics about the
real world. When one moves beyond IIoT
sensor data, many other critical information
models provide context and intent and are
often overlooked.
- 24 -
June 2018