Don't make yourself a data mountain

They say there's nothing like travel for broadening the mind, so I guess my mind must be a little broader having just gotten back from one of those trips of a lifetime — climbing Cerro El Plomo — an 18,000-foot mountain in the Chilean Andes.

When you look around at what's almost a lunar landscape as you get higher, I kept on thinking what a world away it was from my 'day job'.

Now, one thing that's a total pain as you climb are all the loose stones that cover the slopes. They block your path and slow you down so you just long for them to be out of the way.

And it struck me that in one respect, all this loose rubble is like the world of IIoT — and that all these stone pieces are a bit like the mountain of data many companies create for themselves when they set up an IIoT.

?A few weeks before I went away to Chile, I was talking to a potential new customer who had a business that, in many ways, was perfect for an IIoT makeover. And the CEO was certainly extremely enthusiastic about the possibilities. There was no need for me to sell him on the idea of having a highly data-driven business that was going to be much more efficient than his existing model.

But I think we were into our second coffee of the meeting when I had to burst the bubble and refocus his enthusiasm on practical reality. I asked him what was going to happen to all the data that he was going to generate because there was going to be a lot of it streaming off equipment at what was a very large plant.

This is no small issue, but it is one that often gets ignored. But if you're not careful, you can get yourself into a whole heap of unexpected problems.

I could see this happening a few months ago when I was at another firm that was already discussing potential IIoT solutions with another provider. Now, when I heard about what they were proposing, it wasn't so much that what they were suggesting was wrong as such; it was just that they hadn't truly come at things from the customer's perspective. Let's just say they might have gotten a bit overly excited about the size of the customer's budget, which was reflected in a 'bells and whistles' solution that, while covering all the bases, was also going to blow the customer off their feet with a tsunami of information they would be neither able to handle nor know what to do with.

Now, when you implement any IoT system that's going to monitor a manufacturing process, running that system for just a minute or two is going to generate a torrent of new information. And as long as you keep your finger pressed on the switch, then that IIoT system will keep on churning out data, day and night, hour after hour, minute after minute, second after second.

Now, this data needs proper management, analysis, and processing if it's going to mean anything at all. If you don't do that, then it's just noise, like having someone continually shout in your ear but having no idea what they're saying.

So, to me, coming up with an answer that dumps a mountain of data on your customer and then leaves them to fight their way out of it isn't the best call, because in 'solving' one problem, you're making another one.

The whole point of having an IIoT solution is that it should add value to your business, not take it away from it, but if you don't know what all the numbers mean for you or your business, then you're no better off. In fact, you are probably in a worse position because you're stuck with a ton of extra work for little or no reward.

You may have much greater visibility into your operations, but you're actually blind to what's going on.

So, before you start an IIoT project, you need to know the difference between strategically useful data and that which is just tactically generated, the equivalent of all those loose rocks on El Plomo. This is the kind of data that's going to have you sliding all over the place, confusing you about what you should be doing, and leading you down the wrong track when it comes to your IoT implementation. In other words, you've always got to be aware of the potential for misalignment between strategic and tactical. That doesn't mean that you are going to be able to get all the pieces absolutely in sync, but everything has to be facing in the same direction.

In other words, when you bring IIoT on-site along with all the sensors and networking that are required on the plant side, you also need the back-office capability to manipulate data, and for that, you need analysts with specialist expertise and knowledge.

I know of a major compressor company that's got thousands and thousands of machines on site all around the world. Now, their compressors are top-quality and cutting-edge, they can capture every scrap of data being produced. The trouble is that the firm has not yet worked out how to properly leverage all this information to make money from it.?

So, always remember that what you are after isn't data, it's data insight that you are going to need if you are to transform your business.

And as a manufacturer, when you introduce an IIoT system, be ready for the data deluge that comes with it and put in the people and software you need to handle it. And don't turn the tap full on from the start. You need to get into a data Goldilocks zone — not too much or too little, but just right.

If you want to talk about how to get there, then just let me know.