For IndustryWeek’s Smart Manufacturers video series, Senior Editor for Technology Dennis Scimeca visited four different plants to inquire about a specific technology deployment at each. He spoke 5G with John Deere, talked MES development with Hexagon Manufacturing Intelligence, discussed brownfield sites and corporate responsibility with battery manufacturer Viridi and dodged a fleet of robots on the shop floor of an LG plant. This week's podcast episode looks at the series as a whole, and shares key insights gleaned from each video.

Below is an excerpt from the podcast:

Dennis Scimeca: Today I have taken over the Great Question podcast for some shameless self-promotion of our Smart Manufacturers video series. I sincerely hope everyone takes a look at the four episodes. They range between 5 and 10 minutes each and covered private 5G networks, MES development, brownfield redevelopment and social responsibility, and what happens when automated guided vehicles or AGVs and autonomous mobile robots or AMRs pretty much take over your shop floor. 

The first episode covered John Deere's private cellular 5G network that builds off a long history of Deere prioritizing connectivity in its plants. This is Jason Wallin, chief architect at John Deere.

Jason Wallin: When we think about networking historically inside the John Deere plant, there had always been some connectivity, dating back to the 90s, of connectivity between the PLCs and the systems that they needed to connect to, to make very tactical decisions. But based on that, we also knew that we needed more strategic insights and needed more data to be able to collect that. That led to the advanced networking that we see today in our manufacturing environments.

DS: During our conversation with Wallin, we had a concise and excellent rundown about the advantages and disadvantages of networking plants via WiFi.

JW: You know, what is the advantage of using WiFi in networks versus using traditional wired Ethernet? There are really a couple big advantages that I see in using the WiFi over some of the existing Ethernet that we have inside manufacturing plants. 

The first one is the cost associated with the Ethernet is very, very expensive. Secondly, it is very, very inflexible, meaning you can't move it 10 feet to its next location without disassembling that cable. and then installing new cable to be able to support that run. 

Sometimes we've found in the manufacturing environment that that can take weeks or even months to be able to schedule that delivery of those new network drops. Wi-Fi gives you that flexibility to do that. 

But there are some downsides to WiFi as well, meaning that things that move around the shop floor do not transition from access point to access point very seamlessly. We see a big change in latency between devices there or instability. We also know that the density that WiFi can support is not as great as what we can support with the Ethernet technologies.

DS: So how did switching to 5G allow Deere to address these issues?

JW: Why upgrade to 5G technology? I think in our case, there are very good business reasons to be able to do that. 

The first is things that move around the shop floor require that ability to hand off very, very cleanly. Cellular and private 5G give us that opportunity to manage that connection between those devices that moves about the shop floor with a very, very repeatable latency. That's just because of the technology and the algorithms that are built into the protocol and the make before break transition technology that doesn't exist in the WiFi. So that was a very clear case to do that. 

In addition to that, it is also very, very important that we be able to support a high level of density. And in a traditional WiFi access point, we can support about 50 devices effectively. In our microcells, we're able to support 800 devices, and we can support several thousand devices in our small cells that we have connectivity.

DS: Now I find all of that very interesting, but what I found most interesting were the workforce challenges Deere had to face in order to get their 5G network up and running. What do you do when you can't hire the workforce you need to make a tech deployment possible when the labor pool doesn't offer the skill sets you need?

JW: We think about the skill sets that it takes to manage these types of networks in a manufacturing environment, that's a strong understanding of a manufacturing OT environment. It's an enterprise networking focus in how we think about segmentation and division, and then also how we manage the cellular networks. 

That is an almost impossible skill set to find. So because of that, we really found folks that had two out of those three, and then we upskilled them in the final area that they had to be able to build well-rounded individuals that understood all three levels of those expertise. And that's kind of how we dealt with that. That's how we dealt with the fear of the unknown in running our own cellular networks.

DS: So in the end, what did John Deere get for all its trouble? Quite a lot, literally, if we’re talking about data. 

Speaking of data, our second episode covered Hexagon Manufacturing Intelligence's custom Manufacturing Execution System, or MES, as deployed at Hexagon's plant in North Kingstown, Rhode Island, where we shot the episode. And the first questions we wanted to tackle were why anyone needs an MES in the first place and how Hexagon built its own. This is Steve Ilmrud, Vice President at Hexagon Manufacturing Intelligence.

SI: As Industry has progressed and technology has developed there's so much more data available today. It is much more important to have something like an MES that goes out and captures that data, presents it to you in a way that you can action the information. 

A small group of engineers were working in our facility in Germany, and they were developing some software that did automation of data, so moving data from one place to another in our systems. And that became the kernel for what is now our MES today.

DS: MES can get expensive with licenses and updates and customizations. So I was curious how much it costs Hexagon to develop its own software.

SI: By developing our own MES, we, of course, have the upfront cost of having to have a programmer and having the cost of implementation and development. And there are hard costs of the actual developers and soft costs of your workforce, learning the software, learning how to use a system. 

But once those costs are done, those sunk costs are in place, you now are not being held hostage to having to add licenses for every person you add to the system. You've got much better control of your costs going into the future.

DS: I also spoke with Ted Coppa, director of business operations at Hexagon Manufacturing Intelligence, about how constructing your own MES lets you remain agile and expand.

TC: Flexibility is a huge piece in being able to grow into it. But also, if you think you have a best-in-class process or a best-in-class product, is there a likelihood that something off the shelf is going to fit how you do it? 

I don't really think so. I mean, could you shape yourself to somebody else's process? Yes, you most certainly can. But do you want to? Or do you want to build around what's already great about your organization?

DS: What I most strongly took away from this story first was something Ilmrud said about the necessary precursor steps to developing an MES and why you get something out of preparing to adopt the software even if you don't go ahead and implement it.

SI: We had to really understand our processes extremely well. In addition to having the benefit of having the MES in place when you're done, you have now gone through a journey where you've looked at your processes so deeply that you've made improvements. 

So that in itself makes you more efficient, right? The tool will, in the future, continue to allow you to make changes, but the initial impact of getting into your processes, making improvements, is something that anybody can do, with or without an MES system, and is a very valuable exercise.

DS: The second big takeaway for me was the risk of developing your own software, because then you're on the hook for customizations. You can't lay anything on a vendor's update schedule, or priorities.

TC: Did we have to be more sensitive to feedback because we built our MES versus had we purchased one? I think we had to be more responsive. I think you always have to be sensitive to feedback and understanding what that is. 

The difference is when you purchase something, it's a lot easier to say, well, I can't affect that, or it's something else, whereas also a pro and a con is that the answer is “I can affect this. We can do something about it. We're the owner of it.” So I think the sensitivity is the same, but you own the responsibility and the action more so, which makes it more challenging.