The Path from Data to Actionable Information as a Driver for the Industrial Ecosystem
Figure 1: Data and Meaning
observation to enable sensor fusion during
later stages of analysis. Sensor fusion is the
ability to combine many disparate data
sources to form a single stream of related
events and assert causality between those
events.
Without context and intent, one is often
inferring the validity of the outcome. In
highly complex systems, such as
manufacturing processes, it is nearly
impossible to provide valuable insights into
the systems without context. This is often a
key component to the success of any IIoT
solution.
The data cannot be analyzed at this point
since even the shape and structure is
unknown. We can think of Linear A script, as
pictured in Figure 2. If data represented
language, we would be seeing the shapes of
the letters, but not yet know how to
interpret the marks. 6
In Figure 1, we show the process of
transforming data from a raw stream of un-
interpreted bytes to prescriptive wisdom.
The following sections discuss each of these
steps in detail to provide a structured
approach to achieving valuable insights from
data. Many of the examples are
manufacturing-related since that is our area
of expertise.
R AW D ATA
Raw data is collected from the control
systems and sensors to provide continuous
streams of bytes from various sources. At
this stage, it is imperative to create a
consistent timestamp associated with each
6
https://en.wikipedia.org/wiki/Linear_A
IIC Journal of Innovation
- 25 -