Using Metrics in the Industrial IoT Value Chain to Drive Trustworthiness
Figure 5: Example of Resilience reinforcing performance
additional time, complexity and operational
costs as well as initial investment. There is
often a trade-off between safety measures,
for example, and operational agility or
speed. Security can often have a business
impact, in terms of complexity and cost. Of
course, the risks that such trustworthiness
measures help mitigate or prevent, may
have a much larger cost in the long run.
Resilience may take many forms in an IIoT
system. The same system that involves the
above assembly chain may also use an IT
service, say to archive sensor data in a Cloud.
On the IT side the resilience of this service
may be obtained by clustering a large
enough set of servers in different locations
under load balancing, thus mitigating server
failure. That aspect of resilience will also
impact operations in a positive way by
improving requests throughput and
response time. In this case, achieving the
resilience objective by itself does not
improve performance, but the means
deployed for achieving resilience happen to
have a positive impact on performance.
Figure 6 illustrates a negative impact of a
safety measure on some business
performance indicator. For example, safety
may require giving more time to the
personnel to manually change machine
configurations, often needed in high-mix
production. This will require slowing down
the assembly chain thus reducing its
performance. In this example, the
dependency curve shows that increasing
safety will decrease performance (e.g., by
slowing an assembly chain) and vice-versa.
As highlighted earlier in this article,
trustworthiness properties may potentially
adversely affect business performance.
Trustworthiness has investment costs,
raising the question of how much the
organization is willing to pay in terms of
September 2018
- 95 -