Making data smaller is suddenly a big deal
You have to see the Big Picture when tackling Big Data. But that’s no small task.
You can’t buy a Small Mac at McDonald’s. You can’t go into 7-Eleven and purchase a Small Gulp. It’s doubtful the star of “Sesame Street” would have attracted as much attention if he was named Small Bird.
That’s because we love big things. It’s pervasive in our culture. We love big shows and big tops and big valleys. Smaller canyons won’t do; we need a Grand Canyon. And while we have big cities, there is only one Big Apple.
So it stands to reason that if we were going to name the proliferation of information available to plant managers today, we’d come up with the name Big Data. It fits. It matches the size and scope of the information available to plant personnel today, delivered through a variety of devices. And as the data gets bigger, so too do the rooms we build to try and contain it.
When we look at the explosion of data and data delivery today, we see more than just size. The data delivery is relentless. We have alarms to manage and data points to evaluate and information to ponder.
We must do it all while keeping the line up and running. The flow of information and the flow of product must run in parallel, because it is dangerous when they intersect. That’s when bad decisions are made, when processes break down.
So while we call it Big Data, we in manufacturing also are trying to make it smaller. Companies focus their efforts on helping manufacturers slice through the volume of data and the noise of alarms to get at the right information at the right time.
That’s the key, of course. Leaving aside our national preference for bigness, what is most important about data is not its size, but its ability to help us understand the world we live in.
For example, we use temperature to tell us how many layers of clothing to put on each morning. In Chicago over this past winter, the answer was seven. By the time you receive this issue, I’m hoping the answer is down to two. Regardless, I’ll use the data to help me make that choice.
We evaluate data as a matter of course every day. Our new cars can warn us about low tire pressure and high engine temperature, and ping us when we’re running out of gas. While that data may be valuable at various points, what we’re primarily concerned with as we drive is the speed of the car. It’s the essential piece of data we need to effectively operate the car. When we are alerted to other issues, we need to evaluate our need to respond based on where we’re going, how fast we have to get there, and whether this might be a good time to stop and do some maintenance.
Plant managers must make these same calculations, on far more systems and sensors each day. Still, the essential questions must still be answered: where are you going, and how fast do you want to get there?
Synthesizing data is now a crucial part of our manufacturing process. It’s not just about making a decision, because you could respond to any piece of data with a wide range of decisions. Where the large control rooms now play a large role is in bringing all that data to a single point of reference and creating collaboration among stakeholders in the process to understand quickly what the effects of any decision might be on the operation as a whole.
And that’s important because no one person has all the answers. If you create a collaborative process to evaluate systems and solve problems, you bring together all the people who can contribute. If you try to see just one piece of data, or react to each alarm as a crisis, or fail to involve the whole team in the data analysis, you’re going to miss something important.
In other words, you have to see the Big Picture when tackling Big Data. But that’s no small task.
The new control rooms show us that we must do more than just collect and manage data. We must see it as one more tool in our ability to run an efficient plant. We have Big Data, and to a point, that’s a good thing. We have all of the knowledge we could possibly want on what our plant is doing, how well it is operating, and where our equipment’s performance is heading.
We also need a way to shrink that data to fit our needs. Then it becomes Smart Data, and that is a really Big Idea.