The Path from Data to Actionable Information as a Driver for the Industrial Ecosystem
process compared to the actual execution. 15
As one can see, there is a significantly longer
amount of time required to make the part
than predicted, especially concerning the
movement between angular fins (C). If this is
anomalous behavior, the operation can be
flagged as suspect and an analysis can be
performed to determine why the deviation
occurred. From the display, the engineering
estimates are grossly optimistic.
of enrichment is often minutes to hours, but
long-term trend analysis and model creation
are left to the predictive stage.
I NTEGRATION AND E COSYSTEM
The ecosystem integration connects the
enriched and semantic information to the
business systems, allowing for actions to be
taken that have a larger scope than a piece
of equipment and evaluate the impact on
Figure 4: ISA-95 Levels
delivery and revenue. When integrated to
the scheduling and resource management
systems, jobs can be rescheduled to work
around equipment failures or changes to
process plans resulting from feedback from
execution. Data enrichment is still device-
and process-centric when the information is
integrated into the business ecosystem,
systemic changes can be made, repair tickets
One caveat is that at this stage the data is
still point-in-time observations about the
device and the analysis has a very limited
amount of history, only enough to provide
the running statistics that are used for
categorization, signal processing and trend
analysis, as well as comparisons to other
static information models. The time horizon
15
William Bernstein, Thomas Hedberg, Allison Bernard Feeney. 2017. "Toward Knowledge Management for Smart
Manufacturing." Journal of Computing and Information Science in Engineering 17 (3): 23
IIC Journal of Innovation
- 29 -