Always Collect Data

I’ve had a lot of interesting conversations recently around the next wave of the industrial revolution. There’s cool tech with sensors, wireless technologies, 3D printing, machine learning, deep learning, robotics, bio manufacturing, and more.

There is also no shortage of implications: income growth, automation displacing workers while creating new jobs, environmental impacts, water wars of the future, etc. They’re all big topics, worthy of research and discussion.

One of the more manageable conversations about the fourth industrial revolution concerns data collection, specifically when it comes to IIoT (industrial IoT) projects. Working with some of the manufacturing and industrial giants globally, I’ve been lucky enough to hear about some of the newest projects these companies are working on. A topic that comes up repeatedly is agreement on collecting data with sensors. This is unfailingly followed by a prioritization debate over exactly what data to collect.

Inevitably, it goes down the path of “Let’s collect data for which we can foresee a need,” producing a heavy sigh and the urge to face palm myself.

This approach is fundamentally flawed. It assumes there is a known straight line between existing data, questions to ask, and answers to be gleaned. But this approach limits your scope of knowledge to your current frame of reference.

This is where it’s helpful to have a software mindset. If there’s one lesson that IIoT and other projects should take from the world of software, it is that they should collect everything they possibly can.

The data can provide insight and answers that at the start of the project you couldn’t even conceptualize, much less ask. In my less restrained moments, I’ll recommend something like: “If I had a fire hose of all possible types of sensors, I would spray it all over everything, collect and keep it all, and turn loose a bunch of data scientists to frolic in it like ducks. What they will find will blow your mind. Every time.”

Imagine your company builds turbines for a wind manufacturer. You identify areas where you need to collect data based on your current needs. So you collect data concerning failure rate, energy production, optimal conditions for operation, intervals for routine maintenance, and size and weight specifications for turbines in different environments.

When a turbine fails and the crap hits the turbine, as it were, you discover that you did not collect data on service SLAs. You haven’t automated communications to nearby licensed mechanics, and you start missing deadlines.

By collecting all data, no matter how seemingly irrelevant, you will have opportunities to realize increased value from their IIoT projects – preventing downtime, improving deployment of field teams for maintenance, identifying areas to innovate, determining where they shouldn’t waste time, improving efficiencies, minimizing environmental footprint, and more.

Plus, when new technologies replace current ones, you’ll be prepared with baseline analytics.

You will be able to turn chaotic, unmanageable crises into more familiar situations with clearly defined processes.

It’s hard to solve the really big problems when you’re going to cheap out on collecting data that will help you do it.

Next time someone asks “What data should we collect?” here’s the right answer:

Collect All the Data!

Free White Paper: Learn About Standard+Case

Read the white paper

Learn more about taking a holistic approach to IT from Rob England at his blog, The IT Skeptic, and from his books, including Plus! The Standard+Case Approach. For more information, read Rob’s new white paper on turning undefined Case Management situations into new Standard Response Models, written for xMatters: Standard+Case: How IT Response Models Drive Modern Operations.

Request a demo