kentoh - Fotolia
Just as connection and communication in a personal relationship are keys to heading off problems and creating new, more productive behaviors, so it is in the manufacturing realm. At heart, that's just what digital manufacturing is about -- creating connection in service of major improvements in how things are made.
With digital manufacturing, however, we're talking about the myriad processes, machines, people and data on which the creation of anything -- from the simplest product to the most complex piece of machinery -- depends. We're also talking about the seemingly endless technologies that can be a part of the digital manufacturing landscape, including computer-aided design, the internet of things, robotics, 3D printing and more. In other words, the question of just what the term digital manufacturing actually means -- not to mention, why it's important -- can be mind-boggling. To shed light, we turned to Brench Boden, CTO of the Digital Manufacturing and Design Innovation Institute, or DMDII.
A public-private partnership, DMDII focuses on helping businesses build digital manufacturing capabilities and developing a workforce ready for the digital revolution that many experts believe is already happening.
Can you talk about the difference between traditional manufacturing and digital manufacturing?
Brench Boden: Traditional manufacturing has the data but it tends to keep it in the silo; it doesn't try to connect it to the other parts of that lifecycle for better insights. I can give you an example of where that breaks down.
Probably the most oft-spoken-of example typical of manufacturers, even very sophisticated ones, is that the design process produces a design that the manufacturing process -- the factory floor -- can't build. So, the designer doesn't have insight into the limitations of the various equipment and so forth. They design geometry and features and turn it over to manufacturing, and the manufacturing guys have to figure out, 'How do I build this?' [And the manufacturing guys will realize things like,] 'There's no way to machine that feature; there's no way to machine a wall that's this thin without having a quality problem.'
That link between the design part of the process and the manufacturing part of the process tends not to be very well connected, and, in some cases, is incredibly disconnected. That would be one example where making connections between the design process and insight into the manufacturing process is critical for eliminating mistakes and heading off quality problems before they cost a lot of money.
What does that mistake reduction or elimination process ideally look like in this digital manufacturing example?
Brench BodenCTO, DMDII
Boden: You would like to catch manufacturing problems at the design point. So, while the designer -- who's maybe a mechanical engineer, an electrical engineer, somebody who has experience with the computer tools -- while they're making decisions and clicking the mouse and adding features, adding tolerances, you would like to have immediate feedback on the screen somewhere that says, 'Warning, estimate 10% yield from that design feature,' something that flags [an issue] and says, 'Think twice about doing that.' There's a lot of sophistication [needed to provide modelling] in the background to represent the limitations of the manufacturing process in an intelligible way for the designer to use.
Variability happens, and so it's hard to be perfect, but you'd like to minimize mistakes. In some businesses, those producibility problems can be incredibly expensive. If you're producing a large, high-dollar end item and you don't discover you have a problem with the design until you've added a tremendous amount of value and you're deep into the assembly process and you realize this huge major subassembly won't mate with this other huge major subassembly because there is a design problem, it can get really, really expensive.
One of the distinguishing features -- and benefits -- of digital manufacturing seems to be that surfacing of important information throughout the lifecycle to avoid problems or catch them sooner.
Boden: Certainly, avoiding problems is one [benefit], and so we'll … generalize that into better insight at critical decision points. Adding design elements -- those are decisions somebody's making -- [and] having better insight into the ramifications of those decisions is important.
Let's say we are most of the way through production and we have an expensive end item and we realize a serious quality problem: You can't just scrap the item.
So, there's a process, there's a quality engineering process that goes into play for large expensive items, which is: What happened? Has it happened before? Is there approved repair? Is this an operator issue? Equipment limitation? There's this investigation -- almost like the show CSI -- into what caused this quality problem. Well, that means gathering a whole lot of background data. What was the design? What happened with the earlier manufacturing processes? Is this just a bad design so we're going to continue to see this?
So, better decision making later in the process is enabled by having a lot of insight because, every step of the way, you've created a lot of data. Let's hypothetically say we're operating in a digital manufacturing environment, and we've produced some complex product and a lot of our product comes from our supply chain. So, I may have hundreds of suppliers that are providing piece parts and subassemblies to my factory, and then I'm assembling them into this glorious end item -- whatever it is. I want data from those suppliers so that I have insight into the history behind each step of the production process, the tests that they did, the materials they used, and on and on. If I'm a supplier, I may have multiple customers and they may all want different formats when I deliver the data.
I'll give you another example: In heavy duty manufacturing, as in most manufacturing, when you find you have a design issue, you have to make a change to the drawing, electronically, and then you have to make sure that everybody else who's working on that particular part, which may be a huge chunk of your supply chain … has to get notice of that design change. It's called an engineering change notification.
So, that's an electronic paperwork exercise. There's a design change, here's the new revision for this drawing. Every drawing that touches this drawing needs to know what the revision was. It turns out that's really expensive to promulgate through the supply chain -- hideously expensive, in fact. So, again, there's this inefficiency in the way we do things now, even with a lot of connectivity, that, from DMDII's view, suggests there's great business opportunity in ameliorating some of these inefficiencies.
There's a third, broader benefit that most of our member companies are interested in, which is that they have broad objectives for greater agility, the ability to respond to changes in demand and/or desire for customization much more quickly and efficiently; that's a business opportunity that is reflected in market share and customer satisfaction.
Aren't there other benefits to this connected manufacturing, such as traceability and sustainability?
Boden: There's a materials traceability, which goes to environmental concerns. We should be able to facilitate that if we're operating in a digital manufacturing ecosystem. [There is also] the internet of things and that is products that have sensors embedded in them that can relay information back to the original manufacturer about performance, customer satisfaction or whatever. That's really valuable data, especially for durable items, knowing how a product was actually used, let's say a heavy piece of construction equipment from a Caterpillar or John Deere or an aircraft from a Boeing or a Lockheed. Knowing how it was actually used by a customer, that particular tractor or that particular aircraft, and getting real sensor data fed back to your designer and your manufacturing process is unbelievable insight. Now I can look and say, 'Wow, we made this design decision, we made this particular feature, we thought this was going to be really great. But it turned out it broke down all the time and cost my customer a lot of money because we didn't understand how they were going to use that product.'
I look at it from the opportunity for digital manufacturing, which is, 'Bring me that information because it helps me design a better product next time.'
I understand that the digital thread concept is a key component of digital manufacturing as well. Can you talk about that?
Boden: Digital thread is a fairly new concept. It's a central theme for DMDII and here's why: The idea of a digital manufacturing enterprise where I have information at stages across the lifecycle where I'm trying to connect those things. I also have information that I want to get from my supply chain or maybe that I want to send out to my supply chain. How do I manage that? Digital thread is a concept based on acknowledging that I need to organize that data in an intelligent way.
The real issue for digital thread is: Can I be purposeful in organizing all of this data across the lifecycle? That is a very audacious goal, a very difficult set of problems to solve. It will be several years before anybody can say they're operating in a full digital thread environment, and that's OK.
Part two tackles practical steps companies can take to create a digital manufacturing strategy.
Manufacturing gets a digital makeover
Essentials of supply chain sustainability
A look at why 3D printing is growing
Supply chain partners benefit from cloud sharing