 So I wanted to first get our panel introduced. But before I do that, I want to introduce our Dean, Dean Mengqiang, is the Dean, John Edwardson, Dean of Engineering, and he received his BS, MS, and PhD, all from Stanford in 1999, 2000, 2003. Prior to coming to Purdue, he was the Arthur Legrand Dolly Professor of Electrical and Electrical Engineering at Princeton. He also has done a lot of work in the entrepreneurship area, in particular the inaugural chairman of Princeton Entrepreneurship Council. In addition to all of this, he has done a lot of startups and written a bunch of books. And in terms of the kinds of people that we see at this position, he's been perhaps our youngest Dean of College of Engineering in a major university. And in addition to all of this, he still manages to do research. And I should add a little bit about his background. Besides doing all of this, he also likes Coke with the scoop of chocolate ice cream, and I should say original Coke, not the diet Coke. And he's very, very energetic. And in terms of what I have seen, where we are going, or we are certainly onto a new curve. So thank you for all of you to come here. Thank you, Kavik. Indeed, real Coke, no diet Coke. And today's panel, I think, is a fantastic demonstration of real energy, a lot of caffeine intellectually involved here. So, you know, Kavik, thank you for bringing together this wonderful group from leaders in academia to government to private sector. And since I will have to be on the plane when our engineering distinguished lecturer, Neil Grisham-Fair, will be speaking later this afternoon, I also would like to take the pleasure to briefly introduce Neil. And I know Neil for, I don't know, 16 years or so, 15 years or so, and has always been a big fan of intellectual horsepower and the creativity of Neil. So I have to shorten the bio introduction here and just highlight that Neil is the world's leading expert at the interface between the virtual and the physical, between the digital and the physical. And he has founded this Fab Lab with many spin-out, Fab Lab Academy fund, and there are over a thousand, I just learned, Fab Lab around the world today. And Neil, of course, is the director of MIT's Center for Bits and Atoms, CBA, I think that's by design. And his intellectual spectrum is wide and deep. I just want to highlight the two of the many books that he wrote. One is called The Nature of Mathematical Modeling. The other is called The Physics of Information Technology. And these two books are always on my bookshelf, and I seek wise counsel and inspiration intellectually from reading Neil's books all the time. And of course, Neil also teaches one of the most popular courses at MIT, how to build almost everything. I can go on and on, but I just want to say thank you, Neil, for coming Purdue Engineering today. I'll have to watch the video recording of your lecture. And thank you for our all three panelists and wonderful moderator. Back to you. Thank you so much. I have a mic here. So I also wanted to introduce our other panelists. David Roberts is the chief of innovation of an Indiana Economic Development Corporation. Prior to that, he was the president of Battery Innovation Center and also has a lot of experience in intellectual property. And he's a practicing patent attorney. So he received his law degree in Indiana University and also a bachelor's in material science at Lehigh. Sorry about the IU degree. Yeah, I understand. So we also have with us, in addition to David, Jeff. Jeff is also a very entrepreneurial leader here right across town. He graduated with a BS in economics in 2000 from Purdue and is now in the logistics area. He has had prior experience at Caterpillar for several years. And now at Oscar Winsky, he's been there for a few years and has grown the company from a few people to over 70 in several locations. And they do provide logistics services and will be a very nice person to have in our panel in terms of discussing the growing pains as well as influence of technology in our. So let's get started with our basic teams here. So I want to cover three teams in particular. And the main team that I wanted to touch upon is looking at US manufacturing and the changing phase of manufacturing in terms of influence of all other technologies, CI and machine learning and 3D printing. And we have several other areas that impacted including internet of things and robotics and so on. So I wanted to first open it up to the panel in terms of giving your position for five to six minutes. And then we'll dive into it. So do you want to get started? OK. So my hope is this isn't just statements, but we get to disagree. So I'll start by observing, I hate regular Coke and I only drink Diet Coke. So I took with that out of the way. I want to talk to start about three scaling trends and then about three meetings. So the scaling trends, there's one Earth, there's 1,000 cities, there's a million towns, there's roughly a billion people, and there's a trillion things. Just order of magnitude. There were single mainframes. There were thousands of mini computers. There were millions of hobbyist computers. There's billions of personal computers. And there's trillions of internet of things. Computerized manufacturing was invented at MIT in 1952. There was one of those. I started setting up community fab labs, which are like the mini computers of manufacturing. There's thousands of those. We're now developing machines that make machines. So not rapid prototyping, but rapid prototyping of rapid prototyping. Those are like the hobbyist computers that's scaling to millions. We're developing assemblers rather than printers or cutters that can make almost anything from a small set of building blocks that's leading to billions of those. And we're developing self-assembling machines that merge the machines with the materials and that's scaling to trillions. And so in the same way that computing went from 1 to 1,000 to a million to a billion to a trillion, literally, manufacturing is now going through that same scaling. And so we're right at the moment in that scaling curve that looks like when the internet was invented. And it's projecting to anybody being able to make anything. And so the biggest trend in manufacturing is to recognize that the future is for consumers to become producers. And rather than companies producing and shipping, you actually produce in the same way that computing is personal. So those are the trends now for the meetings to ground it. Last week, I was in China at the World Intelligent Manufacturing Summit, which is the leaders of Chinese manufacturing. These are the virtuosos that do spectacular high volume manufacturing. And they're looking at a number of issues and trends in that. So that was one meeting. Then I was in Silicon Valley. One of my students built and runs all the computers at Facebook. One of my students built and runs all the computers at Twitter. One is a senior scientist for Samsung in the Valley. One is leading Google X's new technology. So they're all up and down the Valley, students from my program. And those are billionaires and billion-dollar companies. But both the Chinese event and then the Silicon Valley is all based on producing for consumption. The most interesting meeting was neither of those. Both of those are nominally what today is about. The first meeting in China was about the frontiers of manufacturing. The second one was the frontiers of business creation. But what I want to talk about was the third meeting, which is one I had in Oakland. There's over 1,000 of these fab labs, these mini computers of fabrication, tools to make almost anything. They're doubling every year and a half. And they spawned a fab city initiative. My counterpart in Barcelona became the city planner, the architect of the city. They have a great design sense and 50% youth unemployment. Whole generation can't work. So they're filling the city with digital fabrication tools as urban infrastructure. So the same way the city provides electricity or clean water, it provides the means to make. So if you live in the city, you can produce what you consume. And that's grown into a fab city initiative of cities all around the world joining a Barcelona 40-year countdown to urban self-sufficiency. So instead of products going in and trash going out, data goes in and out, but things stay. And so there was a wonderful meeting in Oakland of the city, the local fab labs and makerspaces and local companies, about how Oakland can develop in a different way from the rest of Silicon Valley. Instead of diverging income and shipping things far away, how it can develop around urban self-sufficiency and consumption becoming creation. And while that may sound a little bit like a stretch, we're 10 years into that scaling and personal computing blew up the traditional computing industry. And so for me, the frontier wasn't the meeting in China, it wasn't the meeting in Silicon Valley, it was the meeting in Oakland around creating an entirely new economy around turning consumption into creation, which we come back to Purdue now kind of completely turns on its head where we are because you're not training engineers to engineer for other people, you're figuring out how to empower everyone to become an engineer. Okay. So I'm Jeff Frost, I am local here. I come from industry, private sector industry and I probably represent best a small to medium-sized company that has very real issues, right? The same issues that we're hoping to see benefit from in the next couple of years, automation in the workplace. We have the same people opportunities that large-scale manufacturers do. There's this gap between where are you at today and then where are you gonna go tomorrow and how does the technology, how does infrastructure help that? For a small company like ours, we're around 350 associates in a company and we're not unlike many companies that would be of that size and we may not have the scale of an Amazon of the world but yet we have to be to the same type of efficiency standards or we don't hold up our end of the deal and we don't exist. So finding scalable solutions for companies such as us is really kind of where I'm at and what I'm trying to figure out how to be better at. So good afternoon, my name is Dave Roberts. The Innovation Officer for the state. I get to serve the state in that capacity. So for any taxpayer here, thank you very much for allowing me to have a job. Appreciate that and Karthik, Dean Chung, thanks for inviting me to take part here. I think today's topic is extremely relevant to the state and to every Hoosier in the state. By per capita metrics, we are the second highest manufacturing intensive state in the nation. Taking away per capita, we're the second highest by GDP in the automotive sector. So within all of manufacturing, a very critical sector to our future. Aerospace defense is also extremely important to us. Let me tie that into then also the people issue that we get to experience. So roughly, actually over 20,000 students graduate each year from our over 50 universities and colleges in this state with a STEM degree, which is a massive asset for the state. But it also turns out that we're the second largest exporter of STEM talent out of our state. So we have a real challenge in retaining you all that are studying engineering, math, sciences and technology here in the state at companies like Oscar Winske and others. So these are the issues that we get to think about. The, a few statements, I don't know if they're gonna be disagreeing, but at least maybe just to set the framework through which we view the opportunities and the challenges that may be controversial. Number one is that apprenticeship has been dying and is dead. People hit the work floor at Oscar Winske with the expectation that they're going to produce and be revenue generating a creative to the bottom line on day one. And that's a real shocker. I know when I hit the floor at my first engineering job with Lockheed Martin, now BAE over in Fort Wayne, I had no clue what I was doing. I was sucking wind and I was losing money for them for the first two years and that's being generous. But you need to be prepared to hit the ground running. And so that's just a reality of where we are. The second thing is that craftsmanship at scale is dead. That's worse than apprenticeship. And then the third thing is maybe tying back to a point that Neil is making just about proximity of manufacturing. If you think about manufacturing, pre-industrial revolution, almost all manufacturing was local. It was within your hometown. If you needed to shoot cobbled, it was down the street. Then you saw centralized manufacturing and then distributed manufacturing so that you could talk to someone like Jeff and think about complex supply chains. And so what we see, I think we're seeing is a return to localized manufacturing, at least for certain components. And maybe that's where we might disagree the most as to which items can be locally manufactured and which ones will not be. Will they all? And what are the challenges for localized manufacturing, for complex or very large structures? Thank you for a nice overview. So the three themes that we will touch upon, the first one related to US competitiveness in long-term manufacturing and how does, how can it look like or the various forms of manufacturing that can come out of such views and vision. The second part we'll touch upon is what David talked about quite a bit, little bit in terms of skills. Preskilling, reskilling. Do we give up on the current generation and start working on the Gen Z because that seems like a transition period? And then the third part is, again, role of governments and institutions like us in terms of STEM training and learning and so on and sort of making them become thinkers and entrepreneurs. So let's kind of get into the first piece of it. Let's talk about productivity. Of course, we have this long-term vision in terms of turning consumers to producers and potentially localization and of course that kind of eliminates competition in some ways. So in terms of productivity, we have seen a lot of changes because of automation and robotics and so on and so forth. But yet, we don't see tangible impact of technology right away in terms of numbers. Some of the long-term technologies like AI take several decades to have an impact. So perhaps we can start talking a little bit about your very long-term vision, maybe 10, 15 years down the road, maybe we're starting to see the beginnings of it and some of the more short-term needs like what you're faced with in terms of logistics. How do we think about how the transitions may happen? Okay, so productivity, excuse me, I think there's three easy mistakes you could make in asking that question. So the first one is to say not much is happening yet. But remember, Gordon Moore in 1965 plotted five data points for transistors on a chip, and that became known as Moore's law. He projected, he noticed that they were doubling, that he projected a decade, it was 50 years. On a linear scale, which is how we perceive, it looked like nothing was happening and then there was a revolution. On a log scale, it was just a straight line. And so exponentials sort of blow up with the vengeance, but it's one, two, four, eight, it looks like nothing's happening, and then boom, it's a revolution. So the thing where you say not much is happening is a straight line on a log plot. It is a revolution happening, it's just not uniformly distributed. Second comment is don't make a false dichotomy between personal and mass production. So software was Microsoft or IBM, open source software came, nobody pays anybody for anything, yippee. Now we have a whole ecology of customizing computers, app creators, app stores, writing for small markets, up to, they're still Microsoft and IBM. Music was the record labels, or you learn to play the piano, Napster came, nobody pays anybody for anything, yippee. That failed, but now there's Spotify, you sell tracks, there's music creation, and so in each of these cases, you fill orders of magnitude of one, ten, a hundred, a thousand, a million. So you mass produce the things where everybody needs the same thing, where there isn't much variation, you locally produce the things that are different. And that has huge economic value. So we've gone off, DEC created the PDPs that made the internet, DEC failed, was bought by Compact, Compact failed, was bought by HP. HP survived because of inkjet printing. Inkjet printing is in Corvallis, Oregon, because they had to hide from their management in Palo Alto because they said we could make beautiful pages cheaply, but it would be slow. Management said, doesn't scale. So they went and hid in the calculator division. But the point is you don't want every page with a low rate printing adds up to a little high rate printing. So the second point is don't make a false dichotomy, rather we're filling in all these scales that didn't exist. But the most important one for productivity is the third one, which is Silicon Valley took years to figure out in the .com crash that Google gives away its product, it gives away search, but it sells as the benefits of searching. And that's true for each of the leading companies. And in the same sense what we see is that the most important product being made in the fab labs isn't selling the thing you make, it's the benefits of making it. So we have fab labs with Alaska natives that have terrible alcohols and unemployment, suicide, but great cultural traditions, and it's transforming their culture to create traditional crafts with modern digital fabrication. We have them at the Protestant Catholic Boundary in Northern Ireland with violent sectarian conflict where people come together to build a community. We're working with the leadership of Bhutan, which you may know is based not on gross domestic product, but on gross national happiness. Doesn't mean they're happy, but their equivalent of the IRS is the GNH office, the Gross National Happiness Office, and what it measures is well-being, health, stress, all these indicators, and what you have to produce is well-being as the product of the economy. It's all over the country where the making improves the well-being. And so the third point, and out of all of this I'd say this is the key one to take away, is the real goal isn't dollars, it's benefit to society. And measuring that is very different than just measuring economic output. And over and over we find the real kind of product from personalizing manufacturing isn't simply more widgets, but it's the well-being that comes from making. So I have the problems, not the answers. And here's real, a real productivity issue that we face every day and I don't think we're unique in this endeavor. We have associates that are going to start working for us right off the street. And there's always going to be a training curve for them and productivity, your first, we did a study on this in our company. The first week, the person is basically rendered useless. You're taking them through training, you have paid for that time, and you have literally got nothing in return for that associate. We're not unique when in this community is not unique when you can take some of the best manufacturers and some of the best logistics firms, and they may have 100% turnover by pure number in any given year. Do the math on that. That is a lot of productivity lost. Where there's an opportunity, and again, I have the problems. I don't have the answers. So this is where I'm looking for help. How do we aid the quickness and speed of making somebody in a process achievable, not in a week but in a day? Karthik, I think, is some interesting things. How do we maybe, for instance, tie in the human capital, which I'm in the middle of, with the mobility of automation that maybe guides them and aids them into being productive in a much quicker pace. That would be a huge, huge win for industry right now. I preemptively say this. I have the problems. I don't have a lot of answers right now. Just one little vignette on that. We did a study, a project with the Department of Defense that has high tech infrastructure and generally low tech people. And so we did an experiment where we taught at Marmac, a group of soldiers, not how to use machines but how to make machines. And so we taught them machine building. And they usually didn't get to do. They were just taught to work the machine, not to create the machine. And it had this huge impact of it was much more interesting for them and it let them see behind what they were using and turned into a lively collaboration. It was very different than the regimented training they had been in. So I think you'll see a trend if Neil has some of the solutions and Jeff has the problems. I get to split the difference and just be confused for the next 45 minutes, I suppose. But let me throw in maybe a fourth dichotomy and that is, how do you square the circle of increased productivity versus employment? I think there's an assumption right now that as we increase productivity via robotics that we're going to see lower employment numbers. Is that true? Is that not true? Where are we going with employment numbers? So on the robotics piece my lab was doing a lot of work on rapid prototyping of rapid prototyping. So how you go to a fab lab and make a fab lab. It's not just 3D printing or laser cutting but more broadly how you can use machines to make more machines. And to do that, I'll talk a little more about this later today, we had to kind of rip up the internals of CAD, CAM, machine control, motion control to integrate them more closely. But what we then realized having done that is what we're really doing is not rapid prototyping but rapid automation. And in turn what that's leading to is a really interesting movement. I had really interesting feedback from people like if you're familiar with them, Andy McAfee and Eric Brinkelson who have written a lot about the rise of robots and robots displacing work. And they noted something they had never considered was rather than robots being done to people, the robots could be done by people. That the robots could augment people if people can create robots. And so the idea of personalizing automation follows immediately after personalizing fabrication and it's a different development path where the robots augment rather than replace because they don't come from outside, they kind of come from inside. So and then I'll let you ask more questions, Karthik, sorry. So if increasing productivity is based on localized manufacturing at even the personal level what does that do to the cost of the product that's being made? Right, so go back to the inkjet example. Inkjet printing is a low cost per page but it doesn't scale and you have one on your desk. Down the hall you have a laser printer for your work group. In the basement you have a high volume printer and then a few buildings over you have a roll to roll printer. The roll to roll printer wins incrementally but it doesn't win if you consider the setup cost of everything for the local production and so in doing that calculation it's a mistake to leap only and right to the incremental cost per thing. You have to consider the whole life cycle of the effort to design, the effort to set up, the effort to run it, the effort to distribute it, the effort to debug it and then there it's a much more nuanced, much more interesting calculation and only in a sense boring things for which needs are completely repetitive justify the high volume end of it. I just wanted to add some comments to it. So as you increase productivity and start producing more and become more efficient and so on there's also a demand response which is more related to the quality of products and the ancillary jobs that these industries might create and so on and so forth. So it's still hard to predict perhaps or project how automation is going to change the nature of jobs. Perhaps we'll talk a little bit more about that but also we are sort of looking at a transition where skill base is sort of becoming more important in this technology chain. So along with improved productivity which may or may not show directly numbers you can argue about what the measures of productivity are and so I wanted to talk a little bit about the types of technology skills that will become more important but perhaps I can also change the question into something more towards what machines are good at versus what people are good at and where is the borderline between the two and does it vary across sectors. It's one question that I want to raise but also to Neil's comment in terms of manufacturing I think we sort of have to look at software and algorithms and electronics and the physical part of products and the digital physical combination as being manufactured not just the physical side of it oftentimes manufacturing we talk about discrete components but that time has gone in my opinion so maybe you can comment to these two points. Yeah and so I have a way to solve your people problem but you won't like it. Generally innovation programs scare me because they prevent innovation and the reason I say that is my lab has spun off billions of dollars in businesses and MIT has spun off trillions of dollars in businesses. The economic output somebody added up falls between India and Russia so MIT matches the economic output of a billion people which is a crazy number and what you should take away from it is not that at any one time MIT has a thousand or so a few thousand people that are smarter than everybody else it's just culture MIT creates an environment where they're productive but the environment it produces is high entropy I describe it as ready fire aim not ready aim fire we don't aim first because if you aim first you only hit what you aim at we get ready then we don't look and the fab labs I started setting up everywhere I'm joined by my wife Lara who's been to many of these in arctic villages or african shanny towns exactly the profile of the same bright invent of people who come to MIT come out of the woodwork but the refugees from schools full of rules and businesses with regimented workflows and they come to the lab as a place where they fit it's high entropy ready fire aim peer to peer learning project based learning and they flourish in the environment so MIT's core technology in the fab labs core technology is how to create these sort of high innovation high entropy creative environments and so I'm sure if I go a few towns over and say you set up one of these labs out of the woodwork would come kids who are considered soldiers in the school and employees who are considered problem employees and it's exactly because they're bright invent of and creative and they don't have outlets in these places and they're exasperated so I mentioned the example of taking the soldiers and not teaching them as workers of the machines but teaching them to understand the machines and sort of empowering them and it was a wonderful experience and so we're surrounded by brain power but we do everything possible to prevent it to limit it and to constrain it and really thinking about the sort of social technology of it's not rocket science it's sort of ready fire aim project based peer driven creative you don't try to settle all goals in advance sort of that whole package but recognizing that's attractive for brain power and the brain power is everywhere if you don't squash it yeah so Jeff and I found out shared some time at Caterpillar and not to call them out in particular but just to use them as an example of a very large manufacturer of durable goods that has a lot of systems and processes in place there is a process for innovation just let that sink in I mean the innovation really doesn't happen until you start breaking the rules or you get rid of the rules and so how can you have a system for it an ABC system for it now ready fire aim is sort of like a recipe but not really it's more of a license and that's what we really need for innovation and I'm thankful Dean Chung and I were together yesterday but he had to leave early before he had to endure my remarks so this won't be a repeat but as an innovation officer I'm thankful because innovation has become a buzzword but the downside of a buzzword is that people lose sight of what the core really is of innovation it's problem and solution that's it we've identified a problem and I'm sure Jeff could give you a list of a dozen that he's enduring today and trying to figure out and then we all just think well what's the solution ABCDEF no bad solutions yet and then we evaluate which one is going to be the best solution for innovation so let me disagree with that in that that works for a very bounded notion of innovation but for me more important than that is traffic and so what I mean by that is like my lab created a hundred million dollar year business that's the dominant safety sensor for the auto industry to controlling airbags that were killing infants and rear-facing child seats and it came out of I was doing a magic trick and they wanted to contact Houdini and what I had developed was a way to channel fields through bodies to make a spirit cabinet so they could contact Houdini and it just happened an NEC executive was coming through my lab and say my goodness wouldn't that work in an automobile so we made it look like a car they took it to the auto trade show and off it went or my lab created the standard for what became RFID tags the reference readers for auto ID, EPC supply chain and that came from we were developing quantum computers we had to program nuclear spin dynamics and the point of these stories it's a mistake to believe you do basic research then applied research as often as not it goes backwards but each of these stories we didn't realize we had a solved a problem those people didn't realize they should be talking to us the real magic came from crossing dissimilar paths and so I would caution against here's the problem let's find a solution that has a role but I would maximize what's usually neglected which is how to cross people who don't think they should be talking to each other to find these kinds of opportunities to make this less of a disagreement what I would say is that what you've described is solution and matching with the problem right it's still the same components it's just which way are you going and it's in the backwards it's the backwards order which is frequently neglected to the point of not even believing it exists but historically over and over it's one of the most important paths and I think that's important because you know you're going to have a medium-sized company like we are and we don't have the opportunity to really experiment a lot right experimentation costs money and you know so but we're very very opportune on what could be the next structure so I think I agree I think that there is creativity that is absolutely lost in probably today's society that can be fostered and potentially provide benefits to folks like myself we just we have to have kind of that solution that comes fairly quickly into play before we can invest too heavily in it. To make that example concrete I just had a really interesting meeting with the sheet metal industry because I have a group of people in my lab who are leaders in an emerging field of the math of folding there's a gorgeous subject of turning origami into a rigorous mathematical subject but it turns out if you apply it to sheet metal it lets you turn 2D shapes into complex 3D forms as a direct right fab process and there what it means is you need to bring together people who do origami these fundamental mathematicians and then with people who do sheet metal and you know I happen to be an environment where those sort of paths cross but generally you wouldn't have a chance to meet people doing origami as an example So I want to break a little bit to the audience to see if there are any questions comments spin-offs from the discussion so far While you're all rousing up the courage to ask a question I just want to point out something about our bias coming into conversations into problem solving until when Neil said sheet metal what kind of factory did you envision and then when he said mathematics what kind of problem solver did you envision just throwing it out there they're not incongruent they need to come together traditional steel worker and a mathematician it challenges maybe our assumptions coming into a lot of these conversations Hello thanks for all your comments I want to spin off a question based on Jeff's situation and for example start-ups start off with a creative way to solve a problem and it's you can define it like in the beginning it's just innovation phase but at a certain point once you kind of know what you want to do then you like when you for example become a medium-sized company then you kind of know the process you know what you need to do and you're kind of optimizing for a single path so what's the right point or at what stages is it even okay to limit innovation versus always having maybe is the right answer always having some part of the organization which is always kind of hey go out there do whatever you want kind of open innovation what do you guys from my side two parts to that first of all it's important to be clear almost without exception every start-up is completely wrong about its business plan and the only common way the most common way for a start-up to fail is to believe in the business plan what they actually make money from is generally remote from anything they thought and you need the agility to realize you are doing something of use but it's not what you thought but then you're absolutely right it begins to scale and in the life of a start-up there's this horrible thing that happens when you get to about 100 people when you have theft sexual abuse you know all kinds of bad stuff starts happening and you suddenly need insurance departments and HR departments and all these things boring companies have you discover the need for them but then what it leads to so something we've seen in the work I do at MIT is we're talking symbiosis of companies who have to deliver on time, on budget, on schedule can't do free play ready, fire, aim, high chaos with critical functions but they need that but conversely the sort of research I do at MIT when we're left to our own resources nothing happens we progress when it's rounded not when we're doing what companies tell us but when we're talking to the companies and so what's been emerging is an overlay is a really interesting boundary where by law companies are supposed to make money and by law we're supposed to not make money so that's a fundamentally symmetry but other than that we create these multi-functional work groups right at the boundary that straddle them the companies don't try to make our culture but we don't try to make their capacity but instead you overlay them right at the interface and when it works well that works really well but it requires the companies letting go control over saying this is exactly what you're going to do and here's the IP rules and all of that and it requires us really engaging in understanding their problems not doing what they say but understanding their issues I agree it's just a couple days ago and Karthik was actually through our operation and to answer your question you never stop trying to innovate the opportunities that you might have or you'll fall behind that's the fear of a small company that we have is that we will fall behind if we don't continue to find a way to do it better you know for instance am I dreaming up a way and do I have 70 of people right most of them are the pure basic work of what needs to be done on a daily basis so that we can hit a P&L at the end of the day do I have people dreaming up a way that a drone could do all of our inventory accounts for instance in our rack system no but would it help me yes it probably really would those are the types of opportunities that I think they absolutely and they should and I hope they do but you have corporations and companies that are much like me that they don't have that going right now inside of their four walls I have an offer the for this fab lab network so if I start a business at MIT it's really easy to handle investment and incubation but with this fab lab network we had to start an educational program to start a nonprofit to support it and we're just now starting an incubator and the incubator is working with partners like Kickstarter or Dragon Innovation or all of them but instead of bringing the innovator to the incubator we're bringing the incubator to them we're creating a network incubator over this lab network and specifically in the fab lab network there's a group of drone developers there's a couple real wizard drone developers that have spun off an initiative to do custom drones to gather data and so I can connect you to this distributed incubator where you would bring the problem of drones in the factory to them invest in them not as a central thing but as this distributed network and then you in turn would be a test case for them and so I'd be happy to make that connection and it's precisely what I'm saying it's sort of you know it's harnessing more of the brain power of the planet thank you it's cool and that's how it works right or hope how it could work right so it's cool thank you so the question coming back to entrepreneurship and startups reminds me of one of my favorite maxims in life that was shared by a you know the great philosopher Mike Tyson so in the early 90s he was being interviewed before a fight before he met Buster Douglas and the interviewer says you know your opponent has the strategy and in round one he's going to kind of run from you and he's going to round two through whatever he's going to body shots and wear you out and then here's how he's going to beat you and Mike said everybody has a plan until they get hit okay so really in your instance entrepreneurship startups but it doesn't matter even in Oscar Wenske and at Caterpillar you need to be nimble and you need to be able to pivot because as was pointed out the company is not what the original business plan said would in order to do that though you can't just have chaos every single day everybody because I've worked with companies that never could get to an MVP because everybody kept over engineering the solution right so you need to have reasonable milestones and checkpoints where you push out a 1.0 and then you iterate from there if you try and engineer perfection from the beginning you'll miss the good somewhere in between so at the early start of the internet I used to think that it's going to democratize innovation and so on but still we see a few companies dominating the whole space and of course we would like to have a lot more of Bill Gates and Jeff Bezos and so on and so forth what is your view on democratization technology and entrepreneurship in terms of what we are seeing today and what it can be so if you believe me that where digital fabrication is scale you don't have to believe me you have to accept digital fabrication is scaled from one to a thousand but then you believe it's going to go million billion trillion it means anybody can make anything not in isolation but networks data travels globally but things so think globally fabricate locally is a way to summarize it to democratize it but for your underlying the question then is if anybody can make anything who's going to be the next Google or Amazon but there may not be one is one of the most interesting answers for Google or Amazon there are economies of scales that benefited from the centralization but think about going to a restaurant if we're going to go out to dinner tonight are you going to take me to a chain fast food restaurant are you going to take me to an interesting local restaurant that specializes in interesting ingredients and you might know the chef and it's your favorite one probably that the economies of scale in restaurants lead to what I was saying in manufacturing the least interesting food the most interesting food is intensely local and it may look much more like that than in a world where bits travel and atoms stay and production is local or first of all I have to often explain to companies who want to become like the IBM of personal fabrication to say if the machines can make the machines it's not such a great business to make the machines but beyond that it could end up looking much more like the local restaurant not Amazon where because it's physical and it's local there aren't the same kind of control points in the central economies of scale it's an assumption to believe that carries over and there's a pretty good argument it may not have got more so far students turning into entrepreneurs but it's sort of back to the future in a sense this is how it used to work but with the profound difference that everything I'm saying rests on global connectivity one person by themselves one lab by themselves can't do it you need the global connectivity for knowledge but you don't need the global connectivity for mass transport but is this valid for trucks and trailers to Jeff in terms of the sizes and scales of products that you deal with I don't know to be real honest you know the I'll jump in I'll help them out no it doesn't work go ahead we're in between right I guess it's a good school of thought between the three of us here because we probably come from very different sectors of the world and I'm in a daily rat race which you know is kind of the world and you know I'm here to I want the help of aided technology and the think tanks that allow us to be better is that make my life better tomorrow I don't know and I don't know if I'm answering your question to be honest but that's kind of where that's where I'm at that's my expertise and that's what we're trying to improve upon I I would caution the no so I can't persuade I have to demonstrate but there's things like there's an interesting fabcar project there's local motors is another version these are all projects doing DIY rapid prototyping of cars there's an open housing movement of people doing digital fabrication rapid prototyping on the scale of the built environment making housing there are a number of these examples that are pushing length scales that you wouldn't think this would apply to yeah and I guess where I go to the know is I think that that sample size is instructive and important to see but in terms of mass adoption of you know companies that are going to use trailers every day printing them in their back office like that's a bridge too far for me mentally so I'd love to see it yeah no I for the companies you sell trailers to I completely agree because that's an example where their needs don't vary so much and there's economies of scale and they benefit but the existential threat coming up behind you is I wrote a piece in Politico recently about the political implications of digital fabrication and so you know let's take this microphone I'm holding this microphone I want to buy it today I would go to a store I would order it it would go to a supply chain in a warehouse in the warehouse things would come on trucks that then maybe came from trains that then came from a harbor that then came from boats that then were shipped in a container that came from a factory in each of those steps there's companies doing the containers on the other hand you know if we had one of these fab labs here that would be like say three times the size of the stage I could easily make this I could do molding and casting for the tooling I could do surface mount electronics for this and if you do that you don't need the truck trailer anymore and so you know one of the most interesting questions would be you can do large scale forming but if I'm making the microphone I don't need the container ship I don't need the container I don't need the truck trailer there's a role for large scale forming but a number of those things that are currently being mass manufactured further up the supply chain you don't need right and so I think for us as a state I mentioned earlier we're the second highest per capita in terms of states relying on manufacturing it's an important thing for us to be thinking about and skating to where the puck is going because if not we get caught flat footed and we have an entire population of Hoosiers that are relying on manufacturing jobs that aren't there anymore it also gives us an opportunity right every threat is an opportunity if you look at it correctly and so it's an opportunity for us to punch way outside of our weight class if we're being strategic and thoughtful and adopting some of these trends a bit earlier than our competitive states and then I would say I'm not qualified to speak on this but the guy that was in my chair before went on to the Department of Commerce working for Wilbur Ross thinking about manufacturing at a global scale it really throws the entire balance of trade and power upside down now actually maybe we haven't really talked about it but maybe the raw material manufacturer is the guy that starts to gain all the power because you've still got to get the material itself in order to print is that fair or am I missing something on that one? Yeah so that's a really interesting point right now among the most sensitive battles are trade tariffs I'm sure it's really complex locally the consequences of the current trade wars in the world I'm describing it doesn't do battle against them it's sort of an end run around them if you can ship data and produce locally the tariffs just don't matter but to your point the vendor I buy electronic components from is Digikey Digikey stocks 500,000 different types of resistor 500,000 different types of capacitor 500,000 different types of connector underlying those are just three material properties conducting, insulating and resistive you're made from a total inventory of 20 parts called amino acids and those 20 parts make up all your complexity and so in the upstream research we're not talking about 3D printer filaments we're talking about assemblers of these basic material properties and it turns out that with about 20 properties like conducting, insulating, semi-conducting magnetic you can assemble those 20 properties to make much of modern technology and so you do need material supply chains but biology doesn't rest on finely divided different minerals it just rests on these 20 properties in the same way the underlying research aiming to get rid of the global dependency in the supply chain is boiling it down to this much smaller set of material properties that you can compose and in a very fundamental way that's actually digitizing the construction of the materials at the heart of the research I think we're kind of reaching I want to give the audience another chance to participate and have more questions come out from all of you I think you briefly went over this but for consumers to produce what level of technical expertise would they need to have? So there's two ways to answer that one is take this device so this device has the capacity just a few years ago of a supercomputer not metaphorically, literally this was a cray computer and if you look at the services on this, what it can do this was state of the art frontiers and the most advanced computing just a few years ago if you look at the use of this device almost anybody can turn it on and operate it you all customize it you pick your icons and your themes within any one application often there's scripting things you do to personalize and customize it then one step in there's things like MIT's app inventor tools that make it easy to get a beginner start in coding then there's tools to support one step in coding all the way up to writing the core operating system and so it's not a true false statement there's again a powers of ten and there's things you can do in a day a week a month a year and so we see the same thing as true in the fab labs there's things you can do in a week a month a year but we consistently take people from beginners into mastering the skills at MIT I teach this very popular class called how to make almost anything where you learn all the skills that I now teach globally and we mix engineers with artists with scientists with community activists but what's interesting is the artists teach the engineers about engineering and the engineers teach the artists about art and it's done the best way you could say it is a lot of education including probably what happens here is just in case you stack up an educational inventory you later draw on this style of learning is just in time you're learning little bits of knowledge as you're mastering and then progressing and so over and over we teach people to do it but it's done not by teaching them everything they need to do something but by teaching as you're progressing so thank you I think it's time to wrap up I wanted to thank the dean and also the school of mechanical engineering let me suggest if we have just a minute maybe closing comments sure yeah so I think first of all thank you all for coming I think the future is bright here in Indiana we've got a lot of statistics that indicate that we're toward the leading edge but we can't rest on our laurels and we need really really smart people like the folks in the crowd to stay here and we need folks like Neil to migrate here I think in closing thanks I think there's harmony I think that there's harmony for many of the things that Neil was working on and speaks of and can greatly enhance the world by which I operate in so I would conjecture very strongly the people like Neil are here but you can't see them and so what you need to do is provide ways for them to come out of the woodwork and the challenge I'd leave you all with is you're teaching engineering to Purdue how do you teach engineering to the whole state thank you so much for all your participation and also I wanted to thank the school of mechanical engineering for helping us pull all this together thanks