Systems Engineering

Design for analytics in IoT product development

Share this post:

In my last blog post, I wrote about a feedback loop in which operational data can be used to improve design—to optimize products, systems or processes, in support of improved business outcomes. The instantiation of this feedback loop is known as the digital thread. In this blog, I’d like to explore the idea of design for analytics in IoT product development a little further.

The Siamese twins of analytics and design

Before IoT, the requirements that drove product design were largely assembled by marketing, which used inputs such as sales results, focus groups, and maybe a “finger in the wind”. The IoT changes this, in that instrumentation can provide data which, after analysis, can provide insight into how a product should be changed to improve performance against particular objectives, such as profitability and customer satisfaction.

We hear of businesses that have changed business models, offering services instead of products to generate new revenue streams. Whenever I hear of these new use cases, I always think, “Wow, that is so clever. Why didn’t I think of that?”

Become the disrupter rather than the disrupted

To be the disrupter rather than being disrupted, companies are instrumenting their products and assets in operation and collecting tons of data, even if they aren’t yet sure how to use the data. I’m reminded of an example I heard last week from Richard Soley, head of the Object Management Group and Industrial Internet Consortium, in his keynote address at the IBM IoT Continuous Engineering Summit in Long Beach, CA. Richard, in his distinctive presentation style, related an episode in which he was discussing with a manufacturer the amount of data they collected—it was millions of data points per second of operation. Richard asked how many of the data points were being used, and the response was “three”. Richard replied, “Oh, one-third, that’s pretty good!” To which the response was, “No. Three.” (Apologies to Richard for blatantly stealing his humor.)

The point is that clearly the IoT either isn’t generating useful data, or the feedback loop isn’t able to utilize that data very effectively. This example demonstrates that the feedback loops don’t just happen automagically—they need to be engineered.   And they need to be engineered with the end result in mind, which is to provide useful data that can be used to optimize objectives, such as business results.

Build data collection into product development

When designing within the context of the IoT, engineers must plan ahead to collect the data that will support the analytics that will provide the insight that will guide engineering decisions that will improve the design (and yes, I meant to write that sentence that way). Maybe this diagram will help.

graphic showing the cycle or continuous engineering product designDesign for analytics is a vital part of IoT product development

Here we see that engineers and developers must design for analytics in IoT product development, which will in turn inform the design (whew, a much shorter sentence!). Of course, this is easier said than done. Success with putting this into practice requires engineers to understand the analytics that will be applied to the operational data and consider how (and when) engineering processes will react to the insights created. Engineers also need to apply systems engineering techniques such as modeling to quantify and characterize this new part of the engineering process.

We’ll talk about some of these principles in my next blog but, in the meantime, you can learn more product development for the IoT by visiting the IBM continuous engineer site.

More Systems Engineering stories

Quick Bytes Live Comes to the Engineering Academy at IoT Exchange

Written by Ryan Boyles | March 27, 2019 | Events, Systems Engineering

Learn from leaders like Walter van der Heiden of Willert Software. Network with peers. Connect with the IBM IoT Engineering community. Get hands-on experience, see real-world results, and prepare yourself to implement innovative solutions that will help you conquer the complexity of your lifecycle management process. more

The Internet of Things requires machine learning and AI

Written by John McDonald | March 20, 2019 | AI Assistants, Systems Engineering

If there ever was a time where there was enough storage to capture it all, that time has passed. It now makes no economic sense to capture every data point, as we need only selected data points to train models. This computer assist of humans through machine learning is almost required to sift through the data from the Internet of Things. more

How artificial intelligence is reshaping the high-tech industry

Written by Grant Bodley | November 29, 2018 | Blog, Electronics

It’s ironic that some companies in the high-tech industry making transformative electronics need to reinvent themselves to have a successful future. They don’t have to start over. But they do need to consider how they will transform their products, services and businesses to answer consumer demand and overcome disruptive market forces. There are two fundamental more