Outcomes, Insights, and Best Practices from IIC Testbeds: Track & Trace Testbed
employ
analytic
experimentation
throughout the stack to decode the data and
apply it in a way that was meaningful for the
enterprise. As more and more data is
collected, data scientists work with the data,
build data models and compute the data.
This complex and ongoing effort of the data
scientists is not something easily
standardized.
model, there are no currently existing
standards.
The Track and Trace Testbed team has
developed a data model which has been
standardized within SAP. Having a common
data model gives flexibility, but it also
enables the standardized structure to be
applied to other current and future use
cases. This standardization has enabled the
testbed team to develop a solution quickly.
An interesting discovery in the midst of this
testbed was the importance of a team
approach: A team must come together
because there is no individual who can solve
all of the issues. Technology integration is
necessary because more than one
technology is essential to fulfill all of the
requirements. So technologies are combined
on the factory level. The team works
together with data scientists, IT experts and
process experts to develop the models to
interpret the data and deliver the business
benefit to the customer. The Track and Trace
Testbed is a joint effort with regular
meetings conducted with IT scientists from
Bosch and SAP, business partners directly
from the customer plant and technology
partners.
For data and communications, there are
many
standards.
The
data
and
communication of the testbed is more
challenging and is still under discussion. The
forklift team may leverage an existing
standard and is currently evaluating which
data and communications standards will be
most useful to the testbed. It is not easy to
figure out the best standardized way to
exchange data between these two worlds:
messages and protocols and within the
technology. Communication standards exist
and the testbed team is looking into
standards
(such
as
MQTT)
for
communications purposes between devices
and the back end, identifying the right
standards to apply for the testbed’s
communications purposes.
Experimenting to Determine Standards
There are many standards related to the
factory tooling aspect of the testbed, such as
CIT, HCT, among others. But a layered
perspectives on API standards is currently
missing.
The Track and Trace Testbed team must
experiment at every stage to determine
which standards to apply, whether an
existing standard need changes and whether
their efforts are relevant for the future. This
is time consuming and requires careful
discussion and evaluation.
When working with standards, there are two
steps: first the team applies the data model
standard to the solution they have
developed. This they sell that solution to the
customer, enabling the customer to achieve
the benefits of the data model. Then the
There are two standards categories at play
within the Track and Trace Testbed: the
common data model and the data in
communication. For the common data
44
June 2017