For Small Manufacturers, The Key To Utilizing Big Data Is Simple: Just Start
Small manufacturers will need to invest in new techniques and technologies, such as big data and other Industry 4.0 solutions, to maintain their competitiveness in the future. This article outlines three steps you can take to start on your data upgrade journey.
Nothing may seem more ho-hum than the ball bearing, the tiny steel spheres first patented way back in 1794, grouped by size and set between circular tracks known as races, each sphere nearly identical to the next.
But as it turns out, in an Industry 4.0 world, even the ball bearing is subject to optimization. With new technology, manufacturers are pinpointing ball bearings that fail with more precision, streamlining the way they select sizes, and anticipating supplier complications with enough time to make cost-saving operational adjustments.
How have we improved upon something so pedestrian? The answer, it turns out, lies in the magic of big data.
Small Manufacturers and the Next Industrial Frontier
Small manufacturers are often scared off by big data. Grasping its potential can seem, at first, like an enormous undertaking. By the time they work through the weeds and get to implementation, some wonder, will they even see a return on investment? And, if they’re already profitable, what’s the point?
Maybe this way of thinking was understandable last year, with the boom in demand. But manufacturers’ recent successes now only provide more reason—and hopefully more capital—to focus on shoring up operations for the future. As my colleague Michael O’Donnell, vice president of operations at MAGNET, told me recently, a more competitive market is well on its way.
“If small companies don’t start to adapt some newer techniques—like big data, censoring, cyber protection, and others under the Industry 4.0 umbrella—I’m afraid the gap will become wider and wider between the haves and the have nots,” O’Donnell said. And then, he put it more bluntly: “Eventually, they’ll be out of business.”
The good news: It’s not nearly as difficult as you may think to get going. Here are three basic steps you can take to kickstart and incrementally scale your data practice—from the basics down to the nitty gritty details, like ball bearings—as your expertise in the area progresses.
1. Start Small—But Start
Real-time data monitoring is the fuel that will power the technology advancements bringing you efficiency, cutting costs, and improving quality. But at this first stage, it’s best not to get too tied up thinking about future use cases.
For now, you’re simply looking to implement, and gain the ability to track, assess, and analyze the performance of your manufacturing machines in—you guessed it—real-time. Spend a couple thousand dollars on laying your data foundation and, as you begin to understand what the data is telling you, you may be surprised at the snowball effect it can have on modernizing your operations.
“If you wait to figure it all out and have the final solution, you’ll never start,” O’Donnell said. “Just do a small thing. And as you start to learn, it’ll accelerate.”
There are organizations that can help you enter this so-called industrial internet of things (IIoT). A good place to start is to plug some numbers into a simple ROI calculator, and begin to understand how your investment in real-time tracking will come back to you. If you’re just looking to get more up to speed on digitization, trade magazines are a great resource. Publications such as Industry Week regularly publish stories that help manufacturers more fluently speak tech.
You can pursue implementation by pulling data off new machines or retrofitting old ones, and then expand your procedures across the floor over time. Manufacturing consultants, such as those in the Manufacturing Extension Partnership network, can help you get started once you’re ready for implementation, allowing you to start keeping tabs on things like vibrations, pressures, temperatures, and cooling rates, all of which you’ll learn to connect to historical patterns of when your machines were performing best or on the fritz. Of course, for the DIY manufacturer, there are suppliers like Losant that will provide you tools to build your own IIoT platform.
Companies that manually track processes on a monthly, weekly, or even daily basis are at a fundamental disadvantage, as they’re only ever looking backward to fix problems. With real-time data monitoring, you’ll be able to assess and fine-tune as you go.
2. Analyze and Act
Once you have real-time data, it’s time to put it to good use.
Manufacturers have historically looked at simple trends related to when machines are or are not working. With live data, you can take things a step further and begin to predict problems before they occur.
Imagine this: a machine you’ve used to cut a certain part has run properly for years without issue. Lately, though, you’ve noticed a change in the real-time vibration data coming off the machine.
What does this mean? Maybe nothing—but it could mean that the machine needs a repair, should be monitored, or should be immediately taken out of service. Because you have loads of historical data to compare to, you notice the change likely signifies a specific maintenance issue you’ve seen in other machines. Instead of waiting for failure, you’re able to schedule maintenance and resolve the issue—without unplanned downtime. And just like that, big data has saved you money.
The trick, of course, is homing in on the right metrics to use as predictors. The more information you collect, the better you’ll be able to predict. Several companies local to us here in Northeast Ohio are starting to explore using Microsoft Power BI to manipulate data for analysis and reporting, with promising early results. For more information, considering seeking out local manufacturing groups, who often host how-to sessions with manufacturers in the area to explain how they’ve used various applications.
Again, if you’re just starting out, these details are a couple steps away. For now, just know that the more you’re able to see maintenance problems coming, the more seamlessly you’ll be able to strategize for them, leading to less unplanned downtime—and more productivity.
3. Take Advantage of Computer Modeling—and Gain Efficiencies
All of this brings us back, finally, to ball bearings.
But first we must talk about the concept of simulation tools called “digital twins.”
In the manufacturing world, we’re at the point where we can recreate an entire warehouse, and its individual physical components, in digital form. These virtual counterparts enable us to get an up-close, digital look at what’s happening inside any given machine, while accurately simulating future performance.
This unlocks a new level of precision. Now, rather than having to inspect each ball bearing individually to ensure it is the perfect size for the job, we can be certain of specifications by referencing a digital twin containing its exact measurements. Rather than send back an entire set of ball bearings due to a single failure, we can reference a part number to the supplier, have that supplier pull up the conditions under which that ball bearing was created, and pinpoint any others in the set that may be at similar risk.
Of course, this one instance is just scratching the surface of what digital twins can do for manufacturers. There are supreme implications for productivity and quality.
If we know the specific details of a given part and its usage history, we can take into account even the smallest defects that part develops and create programs that automatically rewrite themselves to account for imperfections. These tiny adjustments keep machines churning longer and at higher productivity, while producing a consistently higher-quality product. Youngstown, Ohio-based Grale Technologies is a prime example of a company working to use digital twins to improve their machining process.
Because a digital twin can be accessed remotely, we could soon see efficiencies in our technology teams, as well, as resource-strapped small manufacturers centralize technology work off-site. No longer will each small manufacturer require their own IT worker.
Some of these benefits remain in their infancy, even for some of the largest manufacturers in the world. That won’t be the case forever. For small manufacturers who want to realize the promises big data will come to provide, they need only to focus on the here and now. Start along your data journey so that you’re prepared for the practices that will become commonplace—a requirement to stay competitive—in the not-so-distant future.