Steve W. McLaughlin
Tom Kurfess
Audio
Audio & Captions
Transcript
Man: [archival recording] This is the challenge for today— to meet competition and rising costs by moving step-by-step towards automation; continuous automatic production.
Steve McLaughlin: Advanced manufacturing is a quickly evolving area that focuses on products from industries like aerospace and medical devices. It requires the latest technology and high levels of design that are considerably more complex than traditional manufacturing, like auto making and steel manufacturing.
[steam whistle]
[applause]
[marching band music]
Steve: This is The Uncommon Engineer.
♪♪
Man: [archival recording] We’re just absolutely pleased as punch to have you with us. Please say a few words.
[applause]
Steve: Our guest today on The Uncommon Engineer is Professor Tom Kurfess, professor in the Woodruff School of Mechanical Engineering here at Georgia Tech and currently serving as Chief Manufacturing Officer at Oak Ridge National Laboratory in Oak Ridge, Tennessee. Welcome to the program, Tom.
Tom Kurfess: Thank you very much, Steve.
Steve: So we're going to talk today about advanced manufacturing and I think a whole bunch of other things around that. Now, we've been hearing a lot about advanced manufacturing in so many different contexts, and so maybe that's the best place to start is could you share your own view of what advanced manufacturing is and a little bit about what you think it will be in the future?
Tom: Sure. So, you know, manufacturing to me is about making things, and advanced manufacturing is really about, what I would say, is producing high-tech product or producing maybe even low-tech product, but with very advanced — with very advanced manufacturing capabilities. So I noticed in the intro you even said, well, hey, you know, not your traditional manufacturing like automotive, but it turns out, of course, spending a good chunk of time over at BMW, automotive is super advanced. You go to those assembly facilities and you'll see lots of robots, lots of AI running around, vision systems and so forth, so it really it's just incredibly high-tech. And a lot of the same technology is applied over in aerospace. You go to Boeing's facility over in Charleston, where we actually have a lot of work going on, and you'll see a lot of these, a lot of these different types of capabilities. So manufacturing, really the advanced manufacturing, is the nexus for automation controls, materials sensing, you know, you name it, it's all rolled in there. And it's a pretty exciting time because manufacturing is critical to the health of the nation's economy, as we've recently seen.
Steve: And so maybe we can drill down a little bit into maybe some examples before we get too detailed. Can you give some — a little more example — about, let's say, in auto manufacturing, you know, auto manufacturing up until about, say, five years ago, and now what it looks like today and some of the techniques, some of the engineering principles, some of the advanced data stuff, what kinds of things are you seeing in things like auto manufacturing?
Tom: Sure. So I think we could take a look at a number of different perspectives. But one, actually, that really stands out, I was having a conversation — I started out this morning with a meeting with one of my students at seven o'clock this morning, and she was talking about using AI to detect defects, let's say, in paint or on the surface of a body, if you will, and you might think about a scratch or a dent and so forth. But even a few years ago — five, six, seven years ago — if you think about the car and the paint job on a car, you see this what we would call “orange peel,” and it really kind of looks like an orange peel. And it was very hard to quantify that. And you actually had experts that would be looking at that orange peel to see if it was acceptable or not, and these were people. And now what we're really doing is now we're training machines; we're actually using AI to take a look at finishes like the orange peel — of course, they catch scratches and dents and so forth — but even the orange peel to say whether it's good, you know, it's acceptable or it's not acceptable. And then not only — and part of that is, well, now we're going to be much more consistent than maybe five or 10 or 15 human beings that might be distributed globally across the supply of the actual production capabilities for, let's say, a global company like a General Motors or a Ford or BMW. But the other thing is, is now you can actually take the feedback from your AI algorithms, your machine-learning algorithms, and feed them back to the robots in the paint shop that are actually painting the car, so you can actually, in real-time close the loop in terms of what you're getting out of your paint systems.
Steve: And the things that I keep reading about advanced manufacturing stem everywhere from the large, mega-scale kinds of, you know, largest pieces of equipment doing the manufacturing all the way down to the micro and nanoscale. And so can you talk a little bit about that? Is there anything that really connects all of them or is it just so diverse that — I'm kind of looking for common advanced manufacturing ideas, concepts throughout?
Tom: Sure. I think we can take several different tacks on that, that the first thing that I would zero in on is really the digital thread. You know, and that is to say, well, gosh, as we produce something, you want to be measuring it. You want to make sure that it's produced correctly. And then you want to — in fact, what we do is we look at, you know, making a digital passport for a product; let's say it's a turbine blade that we might send over to our partners at Delta Tech Ops down at the Atlanta airport. When that turbine blade shows up at Delta Tech Ops, you want to see it show up with all its production information so that you know that it is a genuine part, that it was produced correctly and what we call born-qualified, or you might also call born-certified, so it's ready to go. But you'd like to see the same thing, let's say, for your microprocessor that your Intel microprocessor or TI microprocessor that you're putting in your computer or maybe your remote-control drone and so forth. So a lot of digital capability there. And it's not just a one-way flow of data coming out with the part, but also we're now tying it into, again, to things like AI-type of algorithms, where we're analyzing what the control knobs were, what the environmental situation was when we produced it, and what the quality was, what the type of part, you know, it's capabilities as they came out, and we're learning what the models are. And again, it doesn't matter whether I'm building a ship, you know, and grinding weld beads on a big ship or I'm all the way down to the microprocessor level and actually doing a wafer fabrication. So all of this links together. And it's on a security basis. It's on a digital information basis. It's on a quality-control and a process-control basis.
Another one, actually, which is a lot of fun is even just think about 3D printing. You do it layer by layer, which is pretty much how you put together a microprocessor; so lots of cool synergies and similarities. And of course, those you can start to link together with a lot of your CAD, your computer aided design or computer aided manufacturing CAD/CAM systems in terms of production operations; so lots of similarities across the board, from big to small to, in fact, gigantic.
Steve: You know, and Tom, someone — and maybe this will trigger something, because I think when the light bulb that went off in my brain, which is pretty dim at times, someone described in a wartime situation where you have an aircraft that's just missing a particular part or a particular part has broken, and it’s sitting on the tarmac for weeks at a time till you get a piece, a very simple replacement part, whereas if you had the ability to rapidly 3D print or rapidly manufacture it and ship it, you would — the cost is almost irrelevant; it's the time.
Tom: What's interesting — and actually this is how our colleagues down at the Warner Robins Air Force Base just south of us, down in just around the Macon, Georgia area — I mean, one of the things they do — and I find this very interesting — so you've got older aircraft where they may not have the exact design for a part. And basically there aren't flight-critical parts, maybe a panel or something. But what I found interesting is they also have plastic machines. Now the part they need is metal, but that does take some time and it's not cheap to do the metal and so forth, so one of the things they do is they may make their first part in plastic, try it out, and they find, oh, that’s not quite the right fit and so forth, and so they modify it until they get it right in plastic and then they go over to the metal printer, and they actually produce it along those lines. So that works really quite well.
The other one that I thought was really, really pretty cool is we were talking to a company, and they make small appliances, and one of the things they said is they said, “Well, can you make molds, let's say, for our injection molars? So when we have a new appliance design, we need just molds.”
And they said, “Typically it takes us about 90 days to get the permanent molds.”
And so making a permanent mold is a little bit harder and heavier duty, you know, and you might want to use the traditional ones, but they said, “Could you come up with a mold that would last 90 days?”
And so, “Well, why do you want it to just last 90 days?”
They said, “Well, once we finalize the design, instead of waiting 90 days to go into production, we could print them out pretty rapidly — those molds — and go into production right away while we're waiting for the final molds to come through.” And I thought that's absolutely brilliant. So now they could be 90 days earlier into the market, which is it's just — it's really critical in this particular area. So, you know, lots of great capabilities there where you would really merge the two and you find a very nice marriage between the two, you know, the more traditional technologies and the newer-generation technologies.
Steve: So, Tom, I know that you wear a couple hats — and at least two, and there's probably many more. You know, in our intro we described — we mentioned a little bit about that you're a professor of mechanical engineering here at Georgia Tech, but you're also currently the chief manufacturing officer at Oak Ridge National Laboratory. So I'm hoping — let's pick one of them so you can talk a little bit about, let's say, the Georgia Tech hat — you know, what kind of research is your lab working on? I know that you were a major force in the creation of our Advanced Manufacturing Pilot Facility off campus, so can you talk a little bit about your work at Georgia Tech and that facility?
Tom: Sure. Yeah, and so, but as you pointed out, it turns out actually, the work at Oak Ridge National Laboratory and the work at Georgia Tech, they're not all that dissimilar. I'm maybe a little bit of a, let's say, a slightly different than — I don't want to say “abnormal,” right, but sort of like a little bit of a different bird in terms of a faculty member because we do a lot of what I would say are a little bit more applied research. We actually go after a lot more hardware and so forth. Clearly, there's some good fundamental stuff. We've got great PhDs going out. They’re going to national labs. They’re going to industry. They're going to academia. But we also work closely with industry. So one of the things that I really have focused on over my career is not just the fundamental aspect of it, but taking those fundamental concepts and helping to scale them and move them forward. And that's exactly what AMPF is all about. AMPF, as you've already said, Steve, is the Advanced Manufacturing Pilot Facility, and that we put together with our partners, in particular Delta Airlines and Boeing. And really, how do we take in that particular facility is really saying how do we take some of the benchtop capabilities that we've developed, in our case at Georgia Tech, and help to scale them up and get them out into production, whether let's say it's a Boeing facility like the Charleston facility where they're assembling the 787 aircraft, or it's Delta Tech Ops, where they're doing complete aircraft rebuild and engine rebuild and so forth and sort of working together. And so really my team — and it is something that you have to get used to because you're delivering some pretty important technology that these companies are now counting on you to deliver. And it's really the same thing over at Oak Ridge National Laboratories. I mean, our objective, our mantra, if you will, is to innovate faster than the competition can copy. And I'm a firm believer in this. And you'll see this around now really getting out even into the federal government. And the idea really is, yes, there are competitors that are going to copy. And you can define competitor as another company, it's another country, and so that's fine, but they're going to copy and that's just the nature of the business. But the reality is, if they're copying, but you stay several years out in front now, then you can very much stay competitive and be very, very profitable. Same thing even happens on cybersecurity. It's not about static and building a wall, and that's it; I'm done. You know that there are always — the competition, so to speak — is always innovating and moving forward, and you have to stay one step ahead of them. And so really, at Oak Ridge, the purview is much broader. My objective there is to identify key technologies that are, let's say, in the nascent stage, so in early stages, maybe at universities or in R&D, at the lower-technology readiness level, so benchtop — identify those that are critical to advanced manufacturing for the United States and then work with industrial partners to scale those technologies up to, I would say, about the pilot plan-level; so really scaling it up. So it's not all that dissimilar from what we're doing at AMPF. Now at Oak Ridge, we've got several hundred thousand square feet and we've got lots of, you know, really large pieces of equipment. At Georgia Tech, it's a little bit smaller, but we're doing the same types of things. In fact, the two fit together really well in terms of designing capabilities, scaling it up, and then really helping to transition it to industry.
Steve: Sounds like you're working directly with a whole bunch of different companies, and you mentioned Delta as part of the Advanced Manufacturing Pilot Facility on campus. Are you allowed to say a little bit about the kind of work that you're doing with them? I mean, it just seems like it's a huge benefit to students to be able to see both sides, the fundamental and theoretical, as well as what might end up in operation. S o can you share a little more about specific projects that you're working on?
Tom: Yeah, sure. I mean, I can identify two really nice projects, one on the digital side and then one on the app-dev side, actually a hybrid side. So I'll start with the with the digital. So, you know, the reality is it's just monitoring of things. So Delta — I think it was our first project — they came to us and they said they have two people, and they spend half of their day, each half of their day, going around all the lubrication tanks and the coolant tanks and so forth in the machine shop. And this is what the first you know, your first job at Delta, you know, if you're sort of the newbie technician, this is what you have to do. It's not the best job in the world. You've got to check the tanks and make sure they're full and the whole bit. And so they said, “Can we develop a just a liquid-level monitoring system?” And, you know, that wasn't too bad to do. But then it turns out, gosh, well, we've got to hook it into their network system. We've got to store the data. We've got to track the data. We have to make them available, you know, maybe even on our smartphone and so forth. So we developed that capability, but then we actually developed a lot of the digital infrastructure including security and so forth that goes with all of that.
And of course, it went from just a simple tank level. Then we were actually working with another company. They were they were interested in — I think it was GM — they were interested. They said, “Well, the level thing is cool. Can you do temperature?”
And by the way, you know, most of these things, you know, you get a little box and sensor and processor, maybe an Arduino, maybe a Raspberry Pi, maybe a little x86 box and so forth. But in general, the most extensive parts of these little sensor boxes, sensor packages, are the box that they go in, not the sensor, not the processor, not the Wi-Fi or Bluetooth and so forth. But the bottom line is every time you add a new sensor, it's just another dollar or two dollars. You've got the infrastructure in place. The other thing that was very cool in terms of working with Delta is initially they said, “Yeah, we'd like to monitor this and that” and so forth, but really what it comes down to is all of a sudden they get it in terms of what the digital architecture looks like. They brought their IT people in. The first thing that happens is no, you can't hook anything up. It's all got to be secure, et cetera, et cetera. But then you work with them because they see the value proposition and really move it forward. So it starts out with sensors and linking them up, and it gets back to how's our process doing? Is it doing the right thing? Are we moving things forward and so forth, which is really nice.
Steve: So I'm really curious to hear more about what's happening in advanced manufacturing at the U.S. government level, because I know that you're involved not just at Oak Ridge, but helping other government agencies, you know, think about some of these concepts and, you know, collaborate across agencies. And, of course, you know, the U.S. government, whether it's the military, whether it's Department of Energy and so many federal agencies support our work, it’d be really great to understand what some of that collaboration might look like and really how our own government is thinking, you know, more about these advanced technologies.
Tom: Sure. And again, what's really interesting is how we're thinking about these advanced technologies is really how do you get them into production so that we're comfortable with them; we know what they're doing. I mean, the difficulty that we have — and let's just — I'll just pick on a bearing. I've spent a lot of time in the bearing industry. I mean, let's say like a needle, just a needle that goes in a needle bearing. This thing, you grind it to kind of a few micro-inches in accuracy. You produce two per second at, you know, pennies, if not less — a piece out of super hard material, It's going to be pretty hard to beat that capability and so forth. So understanding that, bringing it all together, but then also helping not just the big guys. So if you've got the big OEMs or the big prime contractors, but here's the deal — and that is that you've got the tier-one suppliers and they're the ones that supply directly to the primes. But then the two, three, four or five and ultimately, you get down to what I would call the small/medium-sized enterprise, the mom-and-pop shop, and the reality is we have got to make sure that they are moving forward as well, because, you know, your supply chain — and I do like the word “chain” there — is only as strong as its weakest link. And so one of the things we’re really thinking about is how do we move them forward in a secure manner, but also how do we get the information in terms of their production operations and processes and really leverage that information in a secure fashion throughout the entire supply chain.
So, again, I'll give you the example, when you're talking 3D printing, you can have two metal 3D printers sitting side by side, all the same settings, all the same material. And the material characteristics of the part produced in this machine versus this machine are going to be different, and why is that? Because there are some slight differences and so forth, and so you really need to understand it and have the models in the control. Well, if I've got some smaller companies that are using these machines, what I'd love to do is get the data from those machines to understand what they're making so we can certify it, but also so we can get better process control. So it's information going up and also information coming back down. So these are the things that we look at is how do we make sure that this technology is securely deployed throughout the supply chain so that we can get what we need and we can get it in a timely fashion, so you don't have that F-18 sitting on the ground because it just doesn't have a part, you know. Or, you know, that you need a respirator, how do you move that one forward? I mean, we're thinking about that. So this is a great conversation that we've had. So I need respirators. Should we stockpile respirators? I don't like stockpiling respirators because we might not need them for another 20 years. Twenty years from now do you want to have some respirator yanked out of storage after 20 years and plugged into you? I don't think so, right. And so really what you'd like to say, well, maybe we can stockpile the components for the respirator. Well, OK, but maybe there are some components that we could do things like injection molding, so maybe we should store the molds for it. And in fact, the reality is, of course, with 3D printing — and this is some of the stuff we've done at Georgia Tech and we also did it at Oak Ridge — you can print up and machine out with a hybrid system. So it's a combination of printing and machining. You can get those molds ready to go pretty rapidly. So now, instead of a stockpile of physical systems, I've got a digital stockpile, and so that's one of the capabilities you look at. And now I could take my digital designs via simulation and so forth, and I can start to put them together with partners. Another partner that we have that we're working closely with is Autodesk. We can put these together, see how they fit together, see if it's going to be functional, understand what material characteristics we bring together. And then, by the way, you turn around — I mean, the stuff we're doing with Autodesk, you know, they've got generative design. So now, you know, again, you're not taking the human out of the loop. The human still puts the design together, but then you say, OK, generative design, why don't you run a bunch of little variations on this thing and figure out if there's a better approach to it or a stronger approach or a more capable scenario. So just tons of things going out there, and a lot of it is, again, linked back into the digital thread.
Steve: How does the government play a role in helping those small companies better leverage technology so that they aren't the weak link in the chain?
Tom: Yeah, so at Oak Ridge, so part of what we do is we help to identify key technologies and scale them up so that they are production-ready. We’ll do some work in terms of helping out the small/medium-sized enterprises, but really our focus, we’ll bring in our industry partners. Typically they'll have about 80 percent of the engineering capability that could do what they want to do, and we'll supply a very specialized 20 percent others to get it moving along. But then we partner with teams so, for example, the different institutes — the digital institute in Chicago, MxD, or America Makes, the additive institute — we partner with them because their mission, yes, is to help scale up this technology as well, but they also have great connections not only to the big primes and the large companies, but also the small/medium-sized enterprises. So really, we're about partnerships. We provide some of the technical scale-up capability, but then we partner up with our industry partners, with our government partners, public/private partnerships to get it out there. And then, you know, once we scale it up, once we've figured it out and we've helped our partners to get it deployed, then we're on to our next set of concerns and projects and so forth.
So if you come out to our facility up here in the Knoxville area, in the Oak Ridge area, if you come up to our floor — it's a little over 100,000 square feet — you see the equipment that's there today. If you come in two or three years, that equipment will not be here. We actually, we don't even own the equipment. It’s on consignment. You know, once we're done with it, we give it back to our partner company. They'll sell it, and they'll actually bring in the latest piece of equipment. So we're always working on state-of-the-art technology. But it's about scaling it up and getting it out there. That's our job over at the manufacturing demonstration facility.
Steve: What's our secret sauce in the U.S. in terms of bringing manufacturing jobs back to the U.S.?
Tom: Well, so I do think part of it is going to be to leverage the workforce that we have here in the U.S., which is a highly skilled workforce. And we have to not only make sure that we have the next generation workforce, we've got to take the current generation workforce and move them forward as well. And again, I think using technologies like smartphones and so forth and things that people are comfortable with. You know, my mother, she's going to be 82 years old. She bought a new car recently and said, “Well, it’s probably my last new car,” and so forth. So it’s a little bit sad there because she said, “You know, another five, six, seven years, I might stop driving.”
And at first I said, “Well, maybe the car will be self-driving at that point in time. But even if it isn't, on your smartphone you can use Uber. You know how to do that.” And so it really it's a game changer. And, you know, I mean, think about it. For somebody out in production to leverage this thing, it's got way more computing power and memory than what we used to put Neil Armstrong on the moon.
So, you know, so I think that that's going to be part of it. Part of it is really the infrastructure as we start to scale up. I think that the U.S. has seen — I think you could take a look at what's going on with the current situation with COVID, you know, and just that manufacturing is important to the U.S. And it's important to have — I don't think you can say, hey, we're going to have everything over here, but I think that we need to have a stronger industrial base. And the reality is, I think we can do it. As you start to take a look at next-generation machines and chips and so forth, I think you can start building it. We do have a lot of that capability here in the U.S., and so I think the opportunities are pretty good. And even think about vaccine for — I mean, a lot of what we're talking about in terms of security and digital passport and so forth works well for vaccine production as well. So I think that there are good opportunities. There are opportunities to deploy new technologies to make things much more efficient. I think things are very bright.
And the other thing I will point out is this whole bringing manufacturing back. The U.S. is an enormous manufacturer. So China is the number-one manufacturer, but the U.S. is by far and away the second-largest manufacturing economy in the world. So it's already here, and I think it's going to grow and certain things are going to, if you will, come back. But, you know, it's not like we're starting from zero; we have a strong manufacturing base here in the U.S.
Steve: What everyone talks about is, you know, how flat the world is now. And I'm really curious about how advanced manufacturing either contributes to a more global, you know, manufacturing infrastructure, say, for individual products or does it separate it? Does advanced manufacturing, say, in India or advanced manufacturing in South Korea or China give them strategic advantages where there isn't a need to manufacture globally? So looking in your crystal ball, where does it — does it help integrate or does it create greater separation between countries?
Tom: So integrate versus this disintegrate, huh? I would say the — you know, I think it does a little bit of both, depending on your perspective. So certainly, as I've already mentioned, that you've got point-of-assembly manufacturing. So it's very useful that you manufacture a lot of your components close to your assembly plant because then you don't have to worry about supply chain disruptions and so forth, so that works out well. And by the way, the same type of technology approach that would say, OK, we can have more manufacturing in the U.S. might go for India and it might go for China and so forth; so those types of local economies, and maybe it might be a little bit more EU-centric and United States and so forth. But I think the nice thing there is, look, manufacturing just from an economic perspective is a good thing. And so to say, well, hey, we only want to manufacture everything here in the U.S. is probably not good for stability in the world. You really want to have some of that spread out. Simultaneously, what I think is very interesting, is you take a look at Boeing and, for example, their 787, or Lockheed Martin and their F-35, they've got partner companies from around the world, or partner countries in companies from around the world, that are making components for those aircraft, and they come back to the U.S. for final assembly, and so everybody is really winning out there. And if you look at the digital thread, you know, if my, let's say if a certain part of the fuselage is made over in Japan then it's shipped over to Charleston for assembly in the 787 facility, it's very nice that the digital passport for that part of the fuselage comes together. You say, yup, this is good. It'll come together. This is how it all comes together, and you can do this digitally. And now let's say I have a modification to my design. I can modify that design. I can verify it digitally. I can make sure the equipment, whether it's Japan or Korea or Germany or wherever it might be, can produce the part and can produce it accurately. We can verify the process, then ship it in and have it assembled correctly the first time. So I think it really works in both directions. and I think it does make the world a better place. It really does help us to integrate a little bit more. But it also helps to, in my opinion, really get that, what we call “democratizing” advanced manufacturing, getting it out there. And frankly speaking, I think that that really helps keep the middle class strong, which is to me is a very important thing.
Steve: For some of our students out there, high school students and junior high students that are listening and our listeners around the world, people know that mechanical engineering is one of the most popular, about the most popular engineering major right now, and you're a mechanical engineering professor. I know you love teaching students, love being in the classroom. I wonder what are you or what's Georgia Tech doing or how are we educating mechanical engineers today to be ready for the world? The picture that you were painting around advanced manufacturing are really preparing them for that dramatically different world.
Tom: Well, that's a great question, Steve, and I think the reality is it's a combination. So clearly you have to understand the fundamentals. And when those students graduate, you know, they're not going to know everything, but the hope is that they're going to be able to go out and learn what they need to learn as they're going through their career. And that's not just an issue of googling something and figuring it out— this is what Google says and Wikipedia says— but understanding the fundamentals so you can say, “Yes, this makes sense and I can extend it out” and so forth. And that's what I would say is kind of the book smart, the fundamental part. But I think there's also something to hands-on education. So you've got to have a little bit of a feel for building things. I mean, that's what engineers do is we build things. We make things. We design things. So that when you design a mechanism or you design an electromechanical system and you hand it off to somebody to build, that they don't look at it and say, “Well, I just- I can't- I can't build this because it's, you know, it's just outrageous” or “You've made the tolerances too tight” and so forth — so the opportunity to go and just try and build things and move them forward. And so the course I teach is ME2110, which is called Creative Decisions in Design. Actually, it's the second year, so the “2” in 2110 stands for second-year. In mechanical every year we have a design and build course where you design something and you go out and build it.
And in 2110 the students are wiring up motors, driving motors, building machines to compete against each other. Typically they're autonomous; they’re autonomous machines. And they absolutely love it. And in fact, I quite often get in a little bit of trouble because people say, well, you're putting too much work on the students. And in fact, we've tried to reduce the load on the students and so forth. But the problem is that the students are having so much fun. I mean, I can show you pictures of a Saturday evening where the machine shop is packed. Of course, this is all pre-COVID-19. But the students are packed in there, and they're working on the machine tools building their robots and so forth because they're super excited and they're super competitive about it — and this is what drives them. And so to me, it's an issue of both hands-on and also fundamentals so that you basically say, “Yeah, I understand what's going on underneath the hood, but I also understand pragmatically how this all comes together.”
I mean, the thing I really love to do is, you know, I can tell you every student coming out of freshman physics can give you all the equations for riding that bicycle and so forth, but you got to get on the bike and you got to get moving in order for it to work, right? And I actually had one — I remember I was talking to one of my colleagues and this kid said, “Well, hey, how about if instead of moving” — when he was learning how to ride a bike — “instead of move it along, why don’t I figure out how to balance first before I start moving, because it'll be safer to balance, you know, in place?”
But I think if you do the — it sounds good, but if you do the analysis, you know, you’ve got to have those wheels rotating to help stabilize you, right. And so it's a combination of the two things that really makes you a powerful engineer.
Steve: Well, Tom, it's been really great to spend time with you today, and I can't tell you how grateful we are for everything you're doing for Georgia Tech, for Oak Ridge. Obviously, that's a special skill to be able to cross all of those different domains, boundaries, and it's incredibly valuable to our students. And we do hear lots of students talk about 2110, you know, as a tough course, but they remember it and they value it, and I hear them tell stories about that years later. So thanks for everything that you've done in that course and that you've done at Georgia Tech. And take care, and we're looking forward to seeing you back on campus real soon.
Tom: Well, great. I look forward to coming back down to Georgia Tech. And then, by the way, I think if there are any alumni out there, in particular those who have gone through 2110, once we get back online and we get everything going again, you know, we're always looking for judges for 2110. You know, feel free to reach out to me or any of the 2110 staff; you want to come down and judge and show your kids what a, you know, what some real mechanical engineering goes on and get them excited about Georgia Tech and engineering, feel free to contact us. All right, thank you, Steve.
Steve: Thanks, Tom. Take care.
Tom: All right. You, too.
[marching band music]
Geekout
Audio & Captions
Transcript
[big band swing music]
Steve McLaughlin: —a lot of consumers or a lot of public, that's not necessarily, you know, research-oriented. I think a lot of people think about advanced manufacturing in terms of things like 3D printing. And so I'm really curious about where, you know, for folks that have their own 3D printers, you know, where, where, where— what is the connection between advanced manufacturing and 3D printing? And more importantly, you know, 3D printing isn't just about doing stuff at home; it's doing it in very, very secure, high-volume, high-cost, high-scale or large size. Can you talk a little bit about that— about where 3D printing has come from, where it's headed, and how it relates to advanced manufacturing?
Tom Kurfess: Sure. So I think, in fact, that a lot of the 3D printing, I think when you go to the industrial scale and industrial products and applications, a lot of times you're talking metal versus people at home might have the polymer, the plastic printers, which are some of the lower end and, frankly speaking, easier to address. But even on the— with the metal ones— so sometimes they're using powder, these so-called “powder bed” systems. The powder is, of course, they have to be careful because you've got to have respirators on. You don’t want to breathe the powders in. They can get into your skin and so forth. And then actually, I think we know even, you even see this with grain silos, right. There is powder that can be very, very explosive and so forth.
But we're also looking at wires. So if you think about the filament, that your 3D printer at home is putting out— the plastic filament— we're now looking at welding wire. So you can imagine instead of the plastic wire coming out, you take the— you take a steel or a titanium wire and you're putting down with a weld gun, so you're actually building up the weld beam and putting that together. So now we can actually make parts out of steel.
I think the exciting part, and I think what's important here, is not so much that we're making these parts and scaling it up and we're going to make thousands and so forth. I think 3D printing— and actually we actually call it hybrid— so it's really a combination of your classic machine tool with an additive capability on it. But I can put material down, but then I can machine it away as well, so it’s like create your own casting before you go and machine it down. But I think the reality there is if you look at— again, I'll go back to automotive because that's a little bit of a home for me— we've been making cars for well over 100 years, and a lot of the processes have been finely honed, and so when you're looking at making hundreds of thousands of vehicles out of a plant every year, it's pretty hard to beat the processes that are in place. So one of the things actually that I'm excited about is, you know, what happens when a new capability like 3D printing comes along? How do you figure out where that 3D printing goes into the overall value chain and production chain so that you can scale it up? And in fact, I see 3D printing is just an example of that. And, you know, what we're going to do is we're going to figure out how to integrate it both on the design and the manufacturing and the quality control and the whole lifecycle analysis. And once we figure out how to integrate that, when the next hot capability comes along or new technique comes along, we'll have a pathway to integrate that directly into production very rapidly.
The other one, to get back to additive, is looking at rebuilding of turbine blades. So here's the turbine blade, and the tip wears down over time, and so one of the things you'd like to do is actually machine that tip off, add material on, then you have to machine it to blend it down. And so that's one of the things we're looking at.
Now, I’ll tell you some really cool stuff that we've recently put together. And these are things right now that's happening also actually in conjunction with Oak Ridge is as I'm putting material down with my laser— and in this particular case, we’re blowing powder into a laser that's just melting the powder right on the tip of the of the blade— the blade is really thin, and so it's hard to get all that heat out there so, in fact, it starts to heat up. And so now we've got a thermal camera. So, again, back to our digital, we're watching it. If the blade tip gets too hot, then we put a pause in the printing of the material, let it cool down. Otherwise it just kind of melts and flops over. So really tying a lot of this together, and then in the end, we've got an entire digital record of building that blade up so we know we're good to go there.
Steve: And so going into the future, that kind of digital record tied to every single part, no matter how small, allows to, you know, over the lifetime of that part, allows you to integrate and learn— I think that was your point from before— integrate and learn all aspects, not just the physical part. That sounds like that's a good example of that.
Tom: But when, in fact, Steve, so you bring this up, so I'll extend this a little further. So it turns out these turbine blades, when they come out of service, I don't know, thousands of hours of service and so forth, they tend to warp and be bent a little bit and so forth, so they're not particularly the original shape. And so we have to measure that before we can go back and repair it to make sure it's within spec and so forth. But one of the things I'm very interested in is you make this measurement, and now all of a sudden we're starting to build up statistics for these worn blades in terms of how they're warping and twisting and so forth. Now, the reality is, is I don't think I'm going to be able to zero-out that twisting and warping, but I think that we're going to be able to knock out a good chunk of it so that when you build a blade initially, you can set it up so that it doesn't warp and twist as much, which basically means your engine over its lifecycle is going to have much better performance. So not only are we learning about in the process that we're doing, but inspecting the blades coming off of those engines, we're actually learning about what happened to them over their previous life.
Steve: You know, we think about security as it relates to manufacturing, you know, I don't think you're talking about making sure that the physical plant is secure. I don't think you're talking about— maybe you are talking a little bit about espionage or security and safety in the manufacturing plant. It seems like because you mentioned it so many times, that security is now a major piece of not only advanced manufacturing, but all manufacturing. Can you shed more light on when you talk about security, how things have changed and what does that mean today?
Tom: So I think a couple of things. So first of all, I think it's about a third of all the cyberattacks in the U.S. are against manufacturing facilities, really targeting manufacturing facilities. The bulk of those attacks are not so much to, you know, destroy or damage or disrupt, but it's mostly to get a hold of information; so targeting, getting information out of those manufacturing facilities that could provide the competition with information on how to produce things. You know that being said, I think you do want to make sure that certainly people don’t want their equipment to be damaged. You hear, of course, about a lot of ransomware and so forth; it's one thing if your computer gets locked up and all your records are locked away, but what happens when you've got a million-dollar machine tool and that thing gets locked up? And you know, how do you protect that? And I think in particular on a lot of the machines, it's difficult because they really don't want you updating your Windows operating system on the machine tooling. I think we've all updated our operating system and all of the sudden the printer doesn't work. And so the machine tool companies, for example, don't want you to do that update that often because they're not so concerned about the printer, but they're concerned about the rest of the machine not working right.
But then you also have this issue of, well, the product could be tampered with. So if I'm making a turbine blade, let's say I could machine— if I'm machining it, maybe I reduce the amount of coolant, so I actually burn the blade, but you don't see it, but I changed the material characteristics because I've overheated the blade in the machining operation. I could even just reduce the efficiency. So instead of producing a hundred parts per hour, I just dial it down to about 90 parts per hour and nobody would be the wiser and so forth. So but I think there are a whole bunch of things that you have to be careful of. And again, there are just— and there are just situations where people will even just make mistakes. And so how do you make sure that you protect your IP, you protect your equipment, and you ensure that when, whether it's a turbine blade or an airbag or a seat belt, when that comes out, that it is— it's made the way it was supposed to be made and it's going to protect somebody's life or serve accordingly and not really endanger anyone?
[marching band music]
