The IIoT Journey and Technology Adoption (Part 2)

DZone 's Guide to

The IIoT Journey and Technology Adoption (Part 2)

As we continue on our journey to the Industrial Internet, take a look at the Proof of Value concept in action to monitor defective parts.

· IoT Zone ·
Free Resource

In my last blog, I outlined two possible approaches to starting the IIoT Journey — a data-first and an outcome-first approach. Using our fictitious company, ACME, I discussed the importance of the Proof-of-Value concept to help identify desired business outcomes from IIoT investments.

Today, I’m going to define an outcome for ACME and provide you with practical steps you can take to start your digital industrial journey.

Proactively Detect Defective (or Going-to-Be-Defective) Parts During Manufacturing

Every manufacturing process generates a certain percentage of scrap due to manufacturing defects or other defects. Further, typically these defects are only detected at the end or near the end of the manufacturing process.

In this example, using IIoT to detect potentially defective parts early in the process can save time, material, or labor cost — or all three — and could be the starting point of a manufacturing IIoT journey. Starting with something of high potential value that directly impacts the bottom-line will increase the success factor in an IIoT journey.

As a Proof of Value for ACME is defined, we'd recommend developing clear steps, deliverables, and an actionable roadmap in order to show value.

Steps to Take

ACME should start its journey with a data exploration exercise to test various hypotheses, which may have direct or indirect impacts on the outcome. As it goes through this phase, more evidence is surfaced to further focus on key areas or entities within the data ecosystem. It also allows ACME to identify a subset of data that can be leveraged for advanced techniques within data science, machine learning, and predictive analytics. This will quickly weed out subsets, which are extremely less likely to impact the outcome.

For most organizations, the usual next step is to perform analytics modeling and development. In this phase, the learning from data exploration can be used to develop multiple analytical models and predictive models to impact the outcome.

In the context of the outcome mentioned above, the step-by-step approach allows ACME to answer certain important questions.

Among the 40 disparate data sources…

  • Which ones are most critical for impacting the identified outcome?
  • What data sources have good quality data, or are there ways to discard some data that is not useful?
  • What entities and variables are actually necessary to develop analytics solution to drive impact?

By spending time on data exploration exercises and/or analytics modeling, it is possible to quickly test various assumptions, mitigate risks and maximize ROI.  

On the other hand, if ACME decided to integrate all 40 data sources without a goal in mind, it could take much longer to first integrate these sources. And, it’s quite possible ACME would have found that only a small set of data sources would be relevant. Or worse, none of the data would be relevant. Either way, it’s not the result ACME would be looking for.


For an organization like ACME, using the outcome first approach and proving the value up front can help save time, mitigate risk — which comes from uncertainty in data — and reduce initial investment to maximize return.

Most importantly, it lets the organization take the right first step, prove and make a case for outcome-first data-driven decision making, and identify the right mix of products, services and platform to impact outcomes.

analytics, iiot, iot, manufacturing

Published at DZone with permission of Nikhil Gulati , DZone MVB. See the original article here.

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}