Outcomes, Insights, and Best Practices from IIC Testbeds: Intelligent Urban Water Supply Testbed
paying for the connectivity in the business
model.
Enhance the visibility to the
operational states of the pump,
allowing the drilling down to the
details of any pumps that have been
connected in a digital twin model
Provide product lifecycle estimates
Provide
pump
room
video
surveillance and access control
Optimize maintenance and repair
workflow by adding a capability to a
maintenance mobile app to view
real time data from the pump even
before arriving the pump rooms.
As the connectivity and data interoperability
issues are being solved, a third challenge lies
in data analytics. To effectively analyze the
data, it requires large amount of quality data
to build models. It also needs a close
collaboration between the OT and IT sides.
Even when quality data have been
accumulated, it is necessary not only to
identify patterns in the data but also
understand what those patterns mean – the
latter can only come from the OT side as
well. More specifically, one needs to apply
OT domain knowledge to the equipment
data by applying ‘labels’ to the patterns
identified in them and use that to build
machine learning models. This in a sense
requires a close-loop of the data lifecycle
and closely integrated systems where
equipment maintenance and repair records
can be correlated with the machine
operational data.
L ESSONS L EARNED
One challenge the testbed has faced is
interoperability - how to connect to the
heterogeneous pumps made by various
vendors, not just those by WPG, that have
been deployed over decades. Even for the
same model of pumps equipped with the
same model of PLC, the data stored in the
PLC could vary. The ability to adapt to
various data formats is important. Sometime
it may be necessary to swap out the old
controllers from those pumps to enable the
connection to them. Combined with other
capability upgrades, this can generate
revenues by adapting or transforming old
equipment to new technologies.
Another interesting point to bring up is that
through the customer engagement in the
implementation of the testbed, the team has
also learned the features the customers
would be interested in, not just those
planned out by the testbed. For example,
providing visibility to the remote assets
offers substantial initial value. The ability to
adapt to the customer’s need is clearly
important. The testbed is able to do that by
reprioritizing the feature deliveries in order
to deliver the best value to the customers.
There is also the issue of establishing
communication from the IoT gateways that
are connected to the pumps to the system in
the cloud. Pumps are usually deployed in
basements where the wireless signal is weak
if not completely unavailable. Therefore,
arrangements must be made with property
management companies to enable wired
connectivity. There is also the question of
IIC Journal of Innovation
All in all, connectivity to equipment and
collecting data are what we usually first
focus on in an IIoT project. However, we
need to have a longer term vision of how to
- 33 -