While Industry 4.0 is typically thought of in terms of robotic arms, highly automated processes, and “internet of things” (IOT) devices, the next big revolution in manufacturing is also heavily steeped in another emerging technology: big data.
FIND YOUR PATH TO BIG DATA
Start by watching the online video replay of the webinar to get the big picture on big data and to hear how 21st Century Plastics took a “figure it out” approach to implementing its own big data strategies.
To access expertise and help on beginning your own Industry 4.0 journey, read about The Center’s Industry 4.0 Opportunity Assessment. It includes both financial/business questions and process/operations questions and the summary report your company receives highlights opportunities for improvement and a personalized technology implementation plan.
For more information, contact The Center West at 616-301-6247 or via email.
Big data was the main topic at last week’s Industry 4.0 Webinar, hosted by MiBiz and the Michigan Manufacturing Technology Council - West. The third webinar in the larger Industry 4.0 series featured comments from big data experts and manufacturers regarding best practices to incorporate big data into organizations.
As manufacturers continue along the Industry 4.0 path, many are beginning to add data science and analytics, or “big data,” into their operations. If IOT devices are considered the engine of Industry 4.0, big data is the fuel, said Dr. Brent Nowak, the executive director of the applied Medical Device Institute (aMDI) at Grand Valley State University.
“We gather (big data), process it and turn it into the energy behind the global revolution,” Nowak said.
Nowak noted the global value of big data is expected to reach $14.4 trillion in 2021, with 35 percent of companies already collecting data by smart sensor technology. That means manufacturers of all sizes need to jump into big data, or risk being left behind, Nowak said.
“It’s an evolution that you’re not going to be able to stop,” he said of Industry 4.0. “(Big data) is an important factor. People in manufacturing today need to pay attention to it. It doesn’t mean you need to jump into the deep end, but be prepared to get your feet cold and wet.”
But Nowak cautions against companies jumping too far into the deep end and investing in more technology than they can handle.
With vast amounts of data produced daily, companies can easily collect far more information than they can realistically process. There’s also the issue of merging a number of different data formats, Nowak said. Software may collect data from machines in Microsoft Excel, while other data may be recorded with PDF, audio or a number of other different file types. Simply merging those formats into data which can be processed can be an immense challenge for manufacturers.
Overall, Nowak suggests small- to medium-size manufacturers approach big data with the simplest strategy and lightest investment possible. Similar to technology companies working in cutting-edge sectors, it’s better to fail fast and learn, rather than be stuck with a large investment in a misunderstood technology, he said.
“Big data is not just a project, it is a business strategy and I’m recommending that you want to be able to fail,” Nowak said. “If you’re going to hire a team of five PhD researchers to do a big project and invest a lot into infrastructure...that might not be the first place to start your small or medium size manufacturer."
21st Century Plastics Corporation, a Potterville-based plastic injection molding company, has adhered to that light and fast approach to data science. While the company has collected machine data since 1994, it recently began searching for other areas where big data technology may be useful to the organization.
“In 2019, we really started thinking — we have our production really taken care of, but what else can we track, and keep data and records on that may help,” said Doug Sanford, a project and process engineer at 21st Century Plastics. “We were thinking (about) what things would put us… out of business or shut us down.”
Sanford and his team decided to integrate sensors and big data technology into their plastic injection-molding machines to monitor the electrical draw per machine. Sanford opted for the “fail fast, fail cheap method” and purchased some cheap amp meters and created an interface that would collect and catalog data from the machines. 21st Century Plastics hopes to eventually use the data to create a predictive maintenance model. The manufacturer is currently working on integrating other sensors into its vacuum system.
For Sanford, the key to 21st Century Plastics Industry 4.0 journey with big data thus far was the “figure it out” approach the company takes toward problems. Instead of relying on vendors, Sanford and his team do the bulk of research, testing, and problem solving themselves.
For those engineers in positions where they need to sell the benefits of big data to executives, Sanford suggests purchasing some cheap sensors and an off-the-shelf microcontroller and experiment on their own.
“If you really think you need to try something, then maybe buy a $10 Arduino and do something at home and bring it in...plug it in and say look what we can do,” Sanford said. “ I did this in five minutes for 20 bucks. Imagine what we can do with a little more time and a little more resources. Getting your foot in the door and then push through the door and just kind of do it."