There’s a kind of madness that sets in when people are lost, be it in woods, desert, or jungle. Detached from the familiar and even from spatial reality, humans are irrationally prone to keep moving in such circumstances and often, experts say, in circles. Foraying for the first time into the wilderness of IIoT (aka data collection and factory connectivity) will feel much the same for foundries—and while maps and compasses cannot be bought for this journey, one simple principle can help leaders find their way: Stand still and define your purpose. Know your “why.” Identify the problem to be solved before lacing your first boot.

AFS Corporate Member Mercury Marine, a global maker of boat motors based in Fond du Lac, Wisconsin, could write a field guidebook on pioneering through the IT, programmable logic controllers (PLCs), and the connected machine forest. Dave Blondheim, director of Global Operations Excellence and Advanced Manufacturing, helped lead his company’s mission, which gathered serious traction in 2017. After nearly five years of listening, learning, growing, and recovering from the occasional misstep, Mercury Marine has emerged with a four-person “Connected Ops” team that serves not only the company’s two Wisconsin foundry operations but all of its manufacturing plants in Florida, Mexico, and Asia. They’ll add a fifth person this spring.

“Data itself does not solve problems; employees use data to solve problems,” said Blondheim. “People assume, ‘If I collect the data all my problems will be solved,’ and it just doesn’t work that way. We need the context, we need the people to bring meaning to the data and solve the problems.”

A related misconception about a company’s IIoT transformation is that the process can be completed quickly with off-the-shelf software “kits” that enable systems to speak to one another, but no such products exist, Blondheim said. 

He knows this, of course, from first-hand experience—in the absence of silver bullets, he helped form an internal strategy and executed the company’s own solution. A Brunswick-owned company that’s one of three marine-focused corporate divisions, Mercury Marine could have spent a tremendous amount of capital in its connectivity endeavor, but in fact it wasn’t the millions most people might assume. Blondheim estimated the company’s outlay comprised an initial six-figure investment in software, plus annual software license fees of approximately $50,000, as well as salaries for four Connected Ops personnel—not bad, he intimated, for a group serving 4,000 employees at four geographic locations whose marine products are distributed worldwide.

A manufacturer of outboard and inboard boat motors, Mercury is a vertically integrated organization that makes a large percentage of its own components, and hence operates two foundries, a machine shop, paint house, and assembly line at its Fond du Lac campus. There, its aluminum foundry specializes in high-pressure diecast and lost foam castings in Plant 17, while about 1,000 feet away, a second foundry known as Plant 98 produces stainless steel propellers using the investment casting process. On the aluminum side of the business, where more IIoT emphasis has been placed so far, the company puts out about 50 million lbs. of castings per year. 

Making Connections

Blondheim, who has an MBA, a master’s in Industrial Engineering, and a Ph.D. in Systems Engineering, started at Mercury in 2013 and was dismayed to discover fairly slim adoption of data harnessing in the metalcasting industry.

“If I wanted to get information or data out of a machine, it literally meant going down to a machine, getting a USB drive, putting it in there, starting the data recording, going back after a week, pulling it out, and then trying to figure out what happened to address the problem I was trying to solve,” he said. “It was all ad hoc.”

During Blondheim’s first year, the company invested in packages of software that hooked up to its diecast machines and stored data on a network location. Essentially, they created a historical connection to machine serial numbers, which allowed them to track and apply information for quality, maintenance, or other problem solving. 

“That led to not only just process data but then how do we get the number of parts produced by the machine instead of having operators manually track that? How do we get all of this information out? It really started in the aluminum foundry and kept growing.”

The real growth spurt came in 2017 when Mercury Marine President Chris Drees, who was then vice president of operations, attended a technology meeting and caught the IIoT bug. When he returned, his vision turned into the catalyst for connecting all the plants with a seamless flow of data. The timing bisected the company’s frenzy to launch new V8 and V6 engines just a year away, which may have added just a little bit of pressure. Fortunately, Blondheim had already been laying the groundwork within the operational framework, and he helped establish the direction in which Mercury would go.

“The question was, how do we, in the foundry—who create the casting and all the data associated with it—link that to the paint line, which is then linked to the machining line, which is then linked to our assembly line, and then it goes out to our final customer as a serial number?” Blondheim said. “How do we get all of that operational data tied together to help us make better product and deliver excellence to our customer?”

At one point, however, they paused to realize it was not just an IIoT map Mercury was navigating. Beyond connecting sensors to equipment and gathering data, the company sought to tie in its ERP system, quality system, and all other data systems, Blondheim explained. 

“It’s not just connecting machines,” he said. “It’s connecting the whole digital fingerprint of manufacturing, and how do we tie it all together to get that data in front of people to make decisions.”

A Tangle of Languages 

The syntax of IIoT/Industry 4.0 can certainly be confusing because, in fact, there’s no universal definition, Blondheim contends. For some, these terms mean data collection, for others it’s maintenance people wearing virtual reality devices to make repairs. In total, the concepts comprised under Industry 4.0 include:

• Simulation and Computer 
• Modeling
• Virtual, Mixed, and Augmented Realities (VR/MR/AR)
• Additive Manufacturing
• Automation, Robotics, and PLCs
• Sensors, IIoT, and Data Collection
• Systems Integration (Software)
• Data Storage and Cloud
• Cyber Security
• Data Analytics, 
• Machine Learning, and
• Artificial Intelligence 

All the more reason why companies must be narrow and clear on exactly what they want to accomplish. And even then, speaking the same language can be problematic. The original team Drees formed at Mercury Marine was a patchwork of disciplines, including representatives from operations, the IT group, and analytics—he called it the three-legged stool, and the team was tasked with working together as a new breed of cartographers to map out the company’s connected journey. 

“The biggest thing we ran into quickly is that each of these groups of people have really good skill sets, but they don’t know how to talk to each other and understand the entire process,” said Blondheim. “All day long, IT can store information in databases and pull it out of databases, but when we asked them about talking to a PLC or a piece of equipment on the floor, they had no idea at all how to do that. When we went to our controls engineers or electrical engineers who keep the machines up by working on PLCs, they could talk PLCs all day long, but they had no idea how to store that into a database and what that data would look like long term.”

At the onset, it was evident that communication—both within the team and throughout the diverse departments—would be Job No. 1. Ultimately, the company evolved a new entity it called Connected Ops led by Blondheim that more or less dwells in the middle of all the subject matter experts and facilitates communication and execution among many moving parts. 

“So, for example,” he said, “I’ll have a manager who comes up and says, ‘I need to have this information to solve this problem.’ Our Connected Ops team will then go to the controls engineer and figure out what sensor we need to add. We determine where’s it going to be stored in the PLC and how often are we going to be looking for it. And once we get that set up, we then work with IT to say, ‘Okay, we have this data, we want to write it to a database this often, it’s going to get stored at this location. We write a lot of it ourselves and IT, at this point, houses our databases. 

“Once we have that done, we have to figure out how we display that back to the person who requested it,” he continued. “So, how do we create a dashboard from this data that’s now stored in the database? And how often is it refreshed? And what does it look like; what are the visuals? The Connected Ops team talks business, talks PLC, talks databases, and talks visualizations and analytics, and ties it all together.”

Where to Start 

Mercury channels the knowledge from its data collection into three main areas and uses data for specific kinds of decision-making. First is business metrics—the ability to identify how many parts have been made, were machines up or down. These data points are translated to show gaps in production, how to solve problems, and reallocate resources to help get more parts out the door. 

Second, the company harnesses process data that allows managers to focus very intentionally on quality. During casting processes, they can visualize the pressures, the temperatures, the speeds of die casting—all of the aspects of machine functions that affect casting quality. “We use that type of data from a troubleshooting quality perspective,” said Blondheim. “How do we improve our product, and how do we make sure that no bad product is moved from one process to the next.”

Third, Mercury focuses on data that is all about machine performance itself, as opposed to how it impacts product. Managers use data to anticipate and avert machine failure, and understand things like volume of die lubricant used, energy used, and how to make improvements.

Foundries beginning their Industry 4.0 trek have to choose where to put their initial focus, and it boils down to two preferences: ease of entry or acquiring maximum value right out of the gate. Blondheim’s advice? Take the easy door first and start with business metrics. 

Gathering and making use of basics like part counts, uptime, utilization, and even time sheets is typically performed manually, so automating the metrics quickly removes headaches and gives supervisors tools with which to manage better, said Blondheim. It also scales the first hurdle easily to show management some rapid, practical value from data collection. 

Not everyone opts for the easy route, though. And Blondheim concedes that those who start their Industry 4.0 efforts with machine monitoring (versus business metrics) may have more initial challenges but potentially stand to realize greater immediate benefit to the organization, since predictive maintenance can prevent costly and complicated operational halts. Monitoring may not be the easiest, but there are ways to make this path less treacherous, he said. 

“Start small and grow,” Blondheim cautions. “A lot of companies feel like, if they have 10 machines on the floor, that they need to connect all 10 of them immediately. Start with your bottleneck machine and get that one up and running; learn from it and make improvements from there. Don’t try connecting all 10 of them because it’s honestly a very iterative process. You’re going to learn as you connect what you want and what you really don’t want. If you can figure that out on a pilot machine and then scale it, you can go much quicker, I think, than trying to do all 10 of them at first. You’ll be able to tweak each one individually as you learn from it. 

At Mercury Marine, the biggest win of connectivity can be summed up in one word: visibility. 

“It just provided a whole level of operational visibility that we never have had before,” said Blondheim. “And with that came productivity improvements. You can quickly see on the third shift, for example, an operator stopped working at the end of his shift instead of running the machine. You could see it on the dashboards. There was self-correcting and the data helps drive accountability. That visibility was a huge one that led to more productivity. 

“There’s definitely a gain from a quality standpoint,” he added, “of just understanding where my quality is at, where my processes are at; is my process running normal or are there anomalies? Because I’m looking at sensors and temperatures and pressures, it drove us into better understanding part traceability. I can actually track this data through the supply chain or the genealogy of the parts.”

Think About the People

People are the company’s secret weapon for connected success, and, through trial and error, Mercury Marine has learned two game-changing lessons:
(1) There’s a very cultural journey for humans in the IIoT expedition. Resistance to using data almost blindsided Blondheim, who thought managers would instantly embrace new intelligence they were offered. 

“I’m an engineer, so I like all the hard math and science, but I realized the cultural piece of it in this whole Industry. 4.0 journey is just as big a hurdle that you have to get through. 

“Everyone likes to say they want data to solve problems, but then you have longtime employees who have been doing something for years without data—then to actually have it, it sometimes goes against the ‘rules of thumb’ they have developed. You have to know upfront that you’ll have to work through this, helping people with their acceptance of data but also preventing data overload.  

Originally, we would get the data and sort of throw it over the wall, assuming people would start using it. Then we realized that’s not the case. So, we spend a lot more time on just trying to train people on what the data means and how to understand and visualize it.”

(2) Manufacturing in general is missing a skill set—and this awareness has caused Mercury Marine to pivot in its hiring approach.

The missing talent link in manufacturing, says Blondheim, is computer science programming. Most of the time, he said, manufacturers have a mindset of hiring only mechanical or material science engineers, overlooking the computer science engineering talent that’s actually needed for the future of manufacturing. Simply put, operational and business connectivity involves a skill that many manufacturers don’t have, he said. 

“The future is finding people with that cyber-physical expertise and getting them into the industry. We can teach them about foundries or machining or other processes—that’s much easier than hiring someone who knows the process but doesn’t understand the IT and analytics side of it. 

Recognizing the deficiency, Blondheim has hired an IT computer science major who will join his Connected Ops team this May. 

“We need to break the paradigm and bring people into manufacturing who really like data, statistics, math, and programming but have a passion for seeing things being made. 

“Kids coming out of school today, they want to go to Google or Facebook, these big tech companies because they’re doing all of this cool AI and machine learning stuff. But we can do all the same things. In fact, I think it’s even more interesting because I actually have an end product to show for it—not just a click and a 'Like.' 

“We have to get that skill set excited about manufacturing and the foundry and what we do—and then start pulling these people in to get us down this Industry 4.0 road.”  

Click here to view the article in the digital edition of February 2023 Modern Casting.