MILL TALK: What Is Industry 4.0 and How Did We Get Here? with MIT Professor David Hardt

This video has been made possible through the support of our museum members to become a member visit charlesrivermuseum.org join please subscribe to follow us here on youtube and hit like if you enjoyed this video [Music] thank you [Applause] Thank you very much that's going to go back there to that stand okay you're wired up around i'm all set yeah i know you'd rather watch these pictures of the montana rockies but i will go to my presentation now so i want to thank dan for that kind uh introduction and even more so for his kind invitation

to get to talk to you about this i uh full disclosure he mentioned that i Had started this master of engineering degree and we also have started an online version of a part of it which we call the principles of manufacturing i'm going to come to an unmitigated advertisement for that at the end um and force everything to tell you that that's the smartest thing you could ever do but we'll we'll see how we get there but i want to talk about um something that Is a a reality a notion an idea that um is

honestly driving a lot of activity in the commercial world we're still struggling with it a little bit in the academic world and i want to tell you how i think um what we're doing in our educational programs works with this and and part of it then of course is what is it and how did we get there so Um what is and unfortunately our style's a little cramped here i was planning to ask a lot of questions and wait for a lot of answers but um if you haven't if you want to do it just

quick shout out the answer and we'll i'll repeat it what is industry 4.0 hearing nothing then i ask the question what's industry 3.0 That's right it's the one it's the lesser one okay then we have this one what's 2.0 and what's 1.0 well we'll get to that there's actually the you know we we have different definitions here but um uh you can do a lot of different things here there's power there's organization there's there's technology i'm i'm gonna choose this one 1.0 what's that the industrial revolution and you know part of that came i'm going

to pick on the one which had to do with interchangeable parts measurement and i've put in the the springfield armory because if you read the history of that that was a combination of the concept of of interchangeable parts you know machine drawings that were references for these things no more hand filing And and one craftsman making an entire rifle you had people making individual components and different technologies for each of those and in fact if you look at that you found that what was also interesting is the armory was a home for what were essentially

independent contractors who came in and ran these machines and made gun stocks and barrels and other things like that but very very interesting there is a Technology 0.0 that i'm going to show you later and then there's 2.0 what's the first thing comes to your mind when i say industry 2.0 mass production yeah that's the one so i want to show you that this lovely film if you haven't seen it this is a 1936 chevrolet factory and i'm going to show this film twice uh but here i just want you to see yeah this is

Mass production making a lot of the same things that was a a chassis assembly line here's a sheet metal stamping line um one of the things you will notice is there are a lot of people involved there's a lot of and there's a lot of stuff flowing through um you know lots of parts and and other things going on okay so let's let's take that as 2.0 mass production Um we had various forms of power there but in that by the 30s it was primarily electrical and the moving assembly line and and all of the

people pouring over these cars and struggling with things and then there's 3.0 uh that's good yeah just in time lean those actually came later but 3.0 in generally in this what's leading up to 4.0 in most of literature that would be Let's just for lack of a better word call it automation now if you looked at that for that chevrolet plant it kind of looked automated these arms coming out and riveting that frame together but we call that mechanized it was not flexible this was more the era of uh flexible automation and i'm going to

jump ahead actually to the 21st century because Tesla has these beautiful videos and here's a tesla factory putting together the all aluminum model s and you know starting from aluminum going through uh the unwinding process using there's one guy using a laser to cut out the blanks putting it into the press and cranking out these these very precise Difficult to make parts as you'll see here in a second at a rate of is one thing one every six seconds okay so um and this the difference between that this and that chevrolet plant that i showed

is effectively automation if you look really closely there isn't a whole lot of difference between that press that made this Aluminum stamping and the press that made the chevrolet stamping fundamentally there's a huge difference in the technology that's driving it that's controlling it and obviously the robots are doing a lot of the tasks that people were doing okay so now we get to 4.0 what is 4.0 uh i okay those are all so here's my answer everybody remember This movie yeah okay i've got one word for you digitalization i was trying to find one word

that kind of captured it and in my experience and i actually learned that that's a different word from digital digitization which is the one i grew up with but it is basically this idea of capturing data everywhere in the organization Having access to that data everywhere in the organization and using it in in simple terms that's the best that i've been able to come with and then you'll see we're going to have this constellation of technologies that help you do that but i wanted to to mention something i have a one of my most wonderful

graduate students Uh graduated went out and became actually a world famous lean manufacturing consultant and by the way i didn't i didn't know anything about lean manufacturing when the zenobi t learned it himself went out became famous founded a consultant company retired early from all the wealth he gained went off and did good works charitable works got back and said i have to start Getting back into these things and started consulting with local firms near his home in canada and i said to him his name is ed constantine i said ed what do you see

when you go out to these companies is 4.0 and he paused and he said dave the companies i go and see these are small medium manufacturers said they've replaced this this is Handwritten cards where you note down production schedules or you might write down you know the status of a of a process a temperature or pressure or a quality exception or something like that they replaced it with this with a with an ipad and i thought well that's pretty funny that's that's silly and then i thought about this and here's a little exercise Thought exercise

i've gone through so just think about this if if you have the means to do this right now go for it but i think i think that's not necessary so try this write something in a notebook you're writing something on your pad right now okay so let me ask you this not to not to pick on you but i'll be picking on myself you're going To write these notes down and you're going to tear the sheet off and put it somewhere you're going to be able to find it will you be able to read it

once you find it if i don't look at it tonight when i get home right okay and correct it it'll be readable okay perfect what if i want to see it tomorrow i can't do that what if somebody in Japan wants to see it tomorrow afternoon okay yeah facts yeah yeah take a picture and send it through the internet that's cheating um so i thought about this i was sitting in my office talking to my students and said here's 4.0 i'm going to write this down i can't read it i'm going to put it on

my desk you already know i won't be able to find it but if i type it into my Computer or i write it into my phone and i put it into a google doc or one of these other free cloud-based things in that instant in the same time it takes to write on that pad it's now available with perfect accuracy it's digitalized and it's out in the cloud and within milliseconds anybody in the world could could look at it it doesn't sound that Profound maybe maybe we're all used to this i'm still getting used to

this but it's that think about that in a manufacturing context the ability to capture information exactly to be able the ability to transmit it anywhere from this workstation to that workstation to everywhere into the plant to the plant manager to the entire corporation to people Around the world to your customers that's pretty profound and one of the things that i've learned going into looking at some of these smaller companies is when you talk about there's the technology the individual technology which is usually a process or something like that a material transformation and then you talk

about the whole enterprise and this has kind of been the sweep of my career worrying About individual processes and new machines and then the sweep to hey what actually makes money and i'll get to that question later there's a huge amount that's going on in this system called a factory and nobody knows what's going on or when you what you do know that's going on is very sparse information i went out there i got a little bit of data here we know that this always happens that way over there and i think we're a Little

bit behind on production there our quality is pretty good but we don't have great records on it because this is where we kept it okay so this ability to have this precision not just in terms of it being exact but it being immediate and frequent is really the backbone of what 4.0 is all about and it's actually fairly new but i'm going to kind of make the point that it it's i Like everything else everything that's new is really old now i just i threw this up here because um when i threw this up i

was going to say isn't this great but i think it's actually going to fit into my punch line so i was visiting a a nice modern machine shop in emporia kansas this summer and let's put it this way they're making money they were doing well but they could not meet their demand They had a little bit of a quality issue and most of all they said we need to double our space because we can't keep up and so we need more machines and more people but we can't afford the space and unemployment was like a

half percent so there was no way to get people uh and so they bought some technology and this is what i would call 4.0.1 the ipad is 0.0 this is this is As simple as it looks this is actually pretty cool stuff and really one of the first levels of what we would call 4.0 each of these squares represents a fancy new machine tool and on their shop floor and this is simply tell you telling you what's going on on that and it's on a screen out in the factory that's about the size of this

screen right here they did this and all of a sudden their Productivity improved because all of a sudden all the different machinists and the manager of the shop knew exactly what was going on and in a positive way you know it was positive reinforcement so and i have some students starting a machine learning company right now actually they're on leave starting it and this is the first level capture the data and just look at it so That's that you might say that's your first small step on the way to 4.0 and share it that's the

other thing this is this is all out there and by the way we haven't done anything fancy with it we're not we're not it's being recorded but we're not trying to model it we're not trying to do any calculations or we're not doing any Artificial intelligence intelligence on it okay so that that's kind of the that's that's a technological thing how did this whole thing of 4.0 come uh come about um there probably a number of different stories about this but i had one of these students who's now starting the company do a little study

on this and he said Well actually it was started by this project that the journal journal germany german federal ministry started back in way back in 2011 that was executed by the german academy of engineering sciences and you know what did it really mean well it meant cyber physical systems i'm not quite sure i know exactly what that mean but that means basically com computers connected to Machines and and back and forth which is also what we mean by internet of things and by the way i'm so glad i'm a mechanical engineer because we make

things you know this this this wonderful museum is filled with things and the internet has finally caught up with us um and then and then the general idea of smart Factories and and that that's a term we can bandy about all night but here's the interesting thing uh that he found i'm going to go through these kind of quickly so what was the research agenda and without getting i don't want to get too technical but network communication standards 20 25 years ago i was on a visiting committee at the national institute of Standards and technologies

in their manufacturing lab we just come and they talk to you for a couple days what do you think the biggest issue that they were dealing with those days was it was this lack of standards they you know everybody was coming out with this automated equipment like i showed you but this this they had this standard general Motors had their own network standard it was a mess so uh these guys said hey that's that's an issue and if we're gonna start doing all these networks and sharing data then security of it is an important thing

this one's kind of an interesting one and keeps coming up but how do we keep employees current on all this and i'll mention that near the end but here's the one i really like Democratize the proliferation of technology so you're supposed to be focusing on small and medium businesses this was really a movement to have all of those companies kind of play in the big leagues with all the big companies um actually the the slide i wanted is coming later but there's a better punch line but right Now let's just say okay so the idea

was to really democratize things so how do we do that this is now what has evolved as a description this comes from the boston consulting group but i think it's gen got some general acceptance that industry 4.0 is this collection of technologies the ones i've talked about you know internet of things Cloud computing meaning you know everything's distributed everywhere and everybody can get to it big data and big data has always been there we just didn't the ideas behind big data have always been there we just didn't have big data now we do and i'll

explain that in a moment augmented reality system integration cyber security simulation nothing new about that but again it's becoming Ubiquitous autonomous robots we'll see about that additive manufacturing okay we could talk about that too but internet of things cloud computing that sort of thing i i kind of think these up here are the real game changers before i go on though um i was asked by some of my colleagues to point out that 4.0 which came out of germany um actually Has parallels in a lot of different areas and and uh dan mentioned this uh

this uh program that we did back in washington back in 2011 the advanced manufacturing partnership all about looking at ways of organizing and advancing advanced manufacturing technology so japan has a program korea and you've heard about the china 2025 probably so Everybody is is fortunately and i'm glad to say that dan was right about the decline of manufacturing but uh what was mark twain's quote the the rumors of my death are premature the interesting thing is and we had data on this for years i wish i still had it here the net value added manufacturing

value added in the united states hasn't really changed almost in Three or four decades and until just about well and now it's about maybe 10 years ago we also had the highest value added china has now eclipsed that but what has changed is the labor content and the productivity so it's it's declined as a fraction of our economy yes it's declined as a fraction of employment But it's still a huge economic engine but here's the interesting thing now this is back to the democratization don't try to read this but basically these are companies by some

rating system in 2016 how far along they were in becoming a 4.0 company the one at the top is samsung way down here is apple and everyone in between is a mega corporation So the point of this is the small and medium guys have kind of been left behind and that's a that's a issue for another day but it's an important it's an important point okay oh yeah so there's just a little one 77 77 of small manufacturers had no internet of things penetration in 2016. and and another thing that's like what i just showed

you That's your machine tool talking to a to a screen nowadays that that shouldn't quite be like that okay so i promise to say how we got here and we've already talked about this we did there was this thing called 1.0 and 2.0 and 3.0 and here here's my point all of a sudden one day we woke up and we were no longer 3.0 it was 4.0 when was that and the interesting thing is that again Looking at this from an academic point of view it's of course been a slow progression and when i look

back it just every year there's a little bit more of this a little bit more of that so what's the big deal about 4.0 any takers what's that well yes thank you very much that's exactly what it is it turns out it if you wanted to point to one thing it's moore's law The law that we double the speed of as everybody is the double the speed of semiconductors and number of transistors and double the capability every 18 months or something like that yeah yeah well so i have some data on that so here's my

point what's really new about 4.0 oh it's the internet right we use the internet oh now we have pcs everywhere that's What's new uh the web yep it's now we have the web that's what's new that's why we got to 4.0 oh no no it's industrial robots that's what's really new ah yes it's wi-fi that's what's really new you see you see where i'm going with this oh it's artificial neural networks that's the real breakthrough thing okay so um and it's oh sorry nc internet Of things but sorry it's the internet of things that's all

what's new so i did a little bit of history on this i could i could have written some of this myself but what's new okay the internet arpanet is now 50 years old that's when we first had the internet and developed that now you know you can argue how much it was involved there but that was when the first Communications were done over the thing that what became the internet the ibm pc is over the 40 years old the web tim berners-lee kind of established that standard and he not al gore you know came up

with the web uh more than 30 years ago the wi-fi just wi-fi is 23 years old artificial neural networks i thought this one was going to be a surprise because i i actually worked With them a little bit back in the 80s doing some automation work they were pretty crude they're actually dated to 1943 the concept um 3d printing stereo lithography 35 years ago and the last one internet of things it turns out that was established when a coke machine was hooked up to a computer at carnegie mellon university so these Are old technologies right

40 50 years old what's different moore's law moore's law um now we have and and i i would say it's even more than moore's law it's moore's law which has made computers faster and cheaper and all these other things but the other thing is how many people have a computer How many people surf the web and do email and things like that take pictures how many people i don't know how many people do this i just learned how to do this how many people use google photos and search for a face okay you're using artificial

neural networks you're using gigabyte memory you're sending stuff around the world faster than you can believe it how much does it cost you How much does google photos cost you zero how much did your computer cost you a little bit more than zero so now you've gotten to the point where you know you can get on the web for almost nothing you can get a a giga something computer for three hundred dollars you've got the cloud i gave you the the google docs example that's absolutely free and yet you can share documents around the world

You've got the cloud an ipad the thing you're going to use to digitalize everything you can pick those up for 100 bucks if you want to get fancy and buy them a mac it's got to be a couple hundred bucks but it's giga everything for almost zero uh dollars wi-fi everywhere you know it's it's penetrating everywhere as i google photos anybody have a home 3d printer And what did it cost you yeah to 300 bucks by the way one of the big explosions in 3d printing was when we started to do that even though that's

not what's having the impact in industry they were all bottled up with patents and expensive things and all of a sudden people said hey i can make these for 300 bucks and they've exploded in interest and you can buy a nest thermostat And a nest thermostat you know about the nest thermostats these are the ones that talk to the to the wi-fi or you can turn your lights on and off with your iphone welcome to internet of things okay so what happened in that 40 to 50 years is the cost reduction is a thousand plus

times in computers sensors and wireless networks that number's actually very wrong i'm going to show you in just a second The speed increase in others the speed of computing is enormous enormous changes on that easily a thousand times maybe ten thousand times or many orbs of magnitude you know we've heard about the memory thing i love to tell my students about the first computer i worked on which was a little desktop thing from digital equipment and it had eight thousand bytes of memory in it Eight thousand they can't believe it and i said it cost

ten thousand dollars and what i think is really important here is this consumer-friendly ability that's why i was asking we all can use this stuff and we don't have phds in in computer science or network science uh it's been made very very easy so things have plummeted in cost these are the classic things oh and industrial robots have come down Tremendously in cost this is the one i like cost of gigaflops the speed of a computer back here in in 1980 to get 100 gigaflops so that would be like a supercomputer i looked this up

there's a table in this 42 million dollars now down here that same that same cost per gigaflop is three cents so i call that prohibitive and the other One is free you're talking about hardware costs but the other costs have become much more significant you're talking about software software yeah yeah exactly and and that's and that's exactly why in fact i think we had the hardware before we kind of caught up because this approach to what we call universal you know for example a lot of the Standards we're using now have been sort of organically

developed and you know we haven't waited for nist to come up with standards and other things and this this consumer-friendly part has still not penetrated into industry as much as i think it should but it's also i think the key to this problem as i said of democratizing things um so you're absolutely right and i Can't probably you're right because we've gone from from when we had 8k of memory our programs also had a few thousand lines of code and now they have ridiculous numbers and storage same thing okay so what so we have all

these technology stories um and i've been pushing more the information Side of it you know knowing what's going on there are these new technologies like additive manufacturing autonomous robots they're on a continuum additive manufacturing in my opinion has just kind of turned the road from being uh still kind of a an interesting way of doing a prototype or making a specialty part um you know for example bmw came by and said we just want or Sorry was gm said we just want to be able to be able to say we three dinner 3d printed a

part for one of our cadillacs you know something like that but to actually use it in production but then you start to see things like general electric saying we can make a part for our jet engine that will have tremendous improvement in performance of the engine that we couldn't make any other way It will eliminate 20 30 other processes and operations and the supply chains that go with them and on and on and on so it's and it will work the engine will work and now what i found just learned some of this stuff recently

is it it's allowing you not just you think of a 3d printer something that moves around and makes a 3d shape isn't that cool and you generally don't care what it's made of but now To make functional parts that's extremely important and people are exploiting this technology to just not do that but to actually you're making the part and the material at the same time so you can actually custom make the material you can change its properties have mixtures and all types of things like that even to the point where talking to some some people

in the National labs they're talking about being able to make again make parts and materials that you can't make any other way you can only make them by printing them point by point so more to be heard on additive manufacturing one thing additive manufacturing will never do is make body parts like that tesla factory as fast as it's making them so it's it's a new wonderful new technology to make things Uh that we haven't been able to so far and it's opening up a lot of design space so i've talked about all this other stuff

now i want to talk about manufacturing and first thing i want to do is is sort of ask the group what is manufacturing that's right making something from something else something something of value converting a collection of small And potentially less value things into something this is great you guys are even more abstract than we are at mit that's right this is great you're not giving me the answer i wanted you're giving me the right answer yeah yeah it's adding value you're all right okay so what you're supposed to say oh it's it's robots and

machine tools and and and ion etching machines and in other Words and additive manufacturing machines and that sort of thing it's all these cool technologies and i would say yes those are tiny little pieces of the whole thing that come together to add value to convert material and other things like that so this is i this is my segue into defining for you what we call the principles of manufacturing And it is to say what's what's unique about the enterprise of manufacturing that you don't find any other place and if this works i'll be happy

so these are the videos you've already seen and this one up here this is industry 0.0 and by the way this is the oldest video this is the second oldest video and this is from january and what you see here of course you see The assembly plant going on here you see the body stamping here what you'll see here is a gentleman weaving a rough cloth called quar made from the the fiber from a coconut shell and it's a very popular utility fabric around the world it's like like sisal or something like that and woven

but it's made into mats that they're used they're used in particular in india India and that's where that was is to to control erosion and other things like that they also make some decorative things and you can buy place mats at your local um store with rubber on the bottom and they'll have car on the top and we saw those being made in the same factory okay completely different factories right and the different centuries here what's The same in all of these that that really kind of says oh that's manufacturing good now you're playing right

into me thanks yeah assembly okay transition yep things things yep they're machines there there are well sort of there was that's right they had a loom yeah yeah by the way a little side story as we were touring that place over in The corner was a bona fide i'd say 20th century uh automatic loom just covered in dirt and everything like that nobody was using it too hard to maintain so that's the real the real question is to to get to to this what's common about all of these and what's different well we know it

was different right the technology was different we went from hand-powered looms to Mechanized assembly machines and stamping machines with a lot of labor intensity to full automation with everybody just kind of standing back so that's those sort of steps of technology that we've been talking about 1.0 2.0 but what was the same in all of them just looking at the picture they're each producing something how many thumb somethings We'd agree it's more than one it's more than a few and what i what i wanted those videos to really get across and i don't know if

i can get them to run again let me see if this works this one in particular is a great one but um and the the the textile done one is two there's stuff moving all the time one have you ever been into a factory when you didn't see stuff moving And you know there is this this this you know the famous movies of the old river rouge plant you know coal and and steel and iron ore and raw rubber coming in and cars coming out the other side and just continually flowing so there's a continual

flow of material through these technologies to produce that part and they're making a lot of each one and each time you make A new part here each time they stamp that each time they assemble something of the robot every time that guy throws the shuttle across what is it an opportunity to do make a mistake exactly and you will and anybody who's ever done this made more than one thing make something be as careful as you can with your brown and sharp lathe you know And take that micrometer out is everyone going to be the

same yeah not if you have a good micrometer so so you know and and you even know even certainly back in the day here but even even with a modern computer assembled car tesla there is variation okay so what we've really kind of said is is backing up and i'm actually These are things we've extracted from experience in our teaching and and our research but also in our on our curriculum what's common that really says this is what's really different you know there's a everybody knows about the big maker movement that's going on now and

high schools have it junior high schools have it colleges have it everybody should learn How to make something and you can go to a computer and draw something you can 3d print it you can laser cut it you can do all these things and it's great it's getting everybody hands-on it's to make one thing how do you distinguish that from here and one of it is this continual flow of new material through processes in a system of processes i have to read these carefully Combining as you said bringing parts together multiple supplier streams creating a

product with minimal variation i think we all know why and it costs a lot of money you have to buy all this stuff to put through the factory you have to have all this equipment there aren't many other enterprises where you need to put that much capital at risk and you better meet the customer demand or you Can have a warehouse full of stuff that nobody wants and in the end you have to make money doing it you can buy a lot of stuff you can process it okay that's was true of the choir factory

it was true of the chevy factory it's true of the the tesla factory i don't know if they figured out that last part yet at tesla but they're working on it and nowadays Variability of the demand of the time frame is is just huge huge you know you don't have long product cycles like we used to so a lot of variability in like you might say in dimension in quality and variation in time so you you this is what if i take those same words and just emphasize them flow through systems with suppliers and variation

lots of capital Demand cost highly variable just you see a theme here variability flow did i mention variability and and flow so we kind of said you know if we're going to really try and really get at essence of things maybe we we could do something like this so one is you know this idea of things flowing and the other is of things varying you know they don't stay the same i had To do that animation and then you sort of that that's that is that's more been extracted from what we do but i think

it i think it's it's a good basis for our principles and then the success factors if you remember nothing else from tonight flow and variation yeah flow and variation but we actually have found over time this these these next four things Which we call the big four um are really good way to kind of assess anything you're doing in manufacturing whether it's a technology a system or these principles only four for success to measure your success or how you're doing can you think of one does somebody want to buy it yeah that would be demand

yeah cost i heard cost yep productivity repeatability Quality time we're pretty much got it yeah exactly so here's what we do quality rate cost and here's and this kind of gets to does anybody want it flexibility okay so i and i kind of do this in in that order because you could be making stuff really fast at low cost but it's such junk that nobody wants it you could be have quality products made At the right rate at a low cost uh and nobody wants it because you're you've got the wrong product and as we

saw you there are all these famous examples like when apple launches an iphone and they can't make enough of them the rate becomes a real problem when tesla launches a car they can they can't make enough of them and their quality problems But they eventually get to it these are very interesting things to measure against and so my point is that when we think for example about 4.0 or 1.0 or 2.0 how do they affect all these things as we went through these technological stages how did it we hope improve these success factors and we

get at uh my point is as an academic how you know how do we teach this and how do we get at it Oh the i think we already know that so this is this is what this is our curriculum that we used to kind of extract these principles and we do it at really four levels we do it at the process level that's each machine is it is it how does it measure uh on quality cost quality rate cost and flexibility at the process level and we do it by by studying and teaching how

to model And understand flow and variation in a process and by the way that's a really fancy name for statistical process control and some other things we also study sorry how things flow through a factory this is one of the great revelations for me as a mechanical engineer who thought that a factory was a machine a machine just one and then to have some of my colleagues studying What happens when you have a bunch of them together they're re-entrant flows like in a semiconductor plant they one goes faster than the other one of them breaks

down one of them is doing good quality one of them doing bad quality and you've got a target to meet demand out here very interesting very interesting stuff and then Believe it or not this has been up until the the certain the current emphasis on things like ai and machine learning supply chains have been one of the biggest topics because many many companies realize that their ability to manage not just the inventory and materials inside their four walls but outside was as important as anything as anything and One could without oversimplifying too much say there's

there's amazon really understanding your supply chain and then uh flow and variation in your organization and management okay driven by these things i've talked about so um let me just briefly give you an example of what i mean by these uh you know so this is just this is from a lab but this is this is what i mean by flow In a process you're putting material in your your this is embossing but you're putting it in you're taking it out you're putting in and you're taking it out everyone is an opportunity for variation um

so you can you can and this has been a classic thing you cannot have good operating standards and so everybody does things a different way you can have machines that are poorly maintained that are not controlled well Um the big one is the material you bring in is different from day to day even from hour to hour and good old natural variations and then in a factory like this this little simulation here is meant to show you what can happen if you're just happily processing through each of these squares and one of them stops working

and these buffers which are holding the inventory start to fill up and then all of a Sudden the system stops and trying to model that in a probabilistic way turns out to be really important and i would say that our students in our in our professional manufacturing degree mainly come from mechanical engineering and a few other disciplines this is completely new to them and one of the most powerful tools they end up using and i think you can understand That something like a system like this with actually a lot of different things going and this

is about the simplest you ever find if i knew exactly what was going on there at all times which is what 4.0 is about i should be able to manage that system a lot better and so that's one of the big impacts that it's had and then uh so variation in the factory system And then the the supply chain and i don't know if this diagram will work but let's just try it this is sort of the idea of a supply chain i've got you know a factory here feeding into this factory i've got a

factory here going to this warehouse which then is trucked over to this factory another factory there is supplying nuts and bolts some of it comes by by rapid you know by Semi this is coming by a ship you're doing all these different timings and all that kind of stuff you finally get it to the last warehouse where it's shipped to the dealer and your happy customer drives away with the product i don't know if you can see that that but that's me what could possibly go wrong with that Right everything possibly so yeah and this

has become a field day for people who like to do large-scale optimization so i want to use ships because they're cheaper and i can get much more on them but that means for example if i'm in the fashion industry by the time the ship gets to europe from asia the fashion season has changed so i have to be able to predict and things like That or i can i can put it on a transport plane but i'm going to pay a lot more for it you don't know exactly what the customer is going to want

there's uncertainty in the supply you're you're depending on this logistics network it gets quite complicated so again this becomes a really important thing to understand If you're going to have a successful manufacturing enterprise and quite honestly i think a lot of the decline in u.s manufacturing we know it came because we weren't paying attention to these things and let me tell you real quickly before i get the last one i just a quick aside and this is where i can bring in my history i've been at mit for 40 plus years and about 25 30

years ago we started This thing called the leaders for manufacturing program anybody ever heard of this yeah you've heard of it do you know why we did it we were scared to death that the japanese were going to completely take over manufacturing and didn't know what to do about it we didn't know what to do at mit and major corporations in the u.s didn't know what to do And so we started this program they put a lot of money into it and it's been a tremendous success now now it's just about manufacturing excellence and and

you know has broadened but it's had a big impact here's the thing the first 10 years or 5 to 10 years what do you think we focused on think about japan think about think toyota think chevrolet Quality yeah so it was all about quality all of our projects and all of the research on quality then what came next think toyota again just in time lean manufacturing that's the system the first was the processes this is the system flowing things through the system yeah yeah well there's a longer discussion around just in time and so it's

more a matter of what what lean manufacturing taught us Is that running your factory really well was important i'll just put it that way just in time was one of many things that were involved and not not a panacea doesn't always work for that very reason um in fact i was in japan once and they there was some sort of a storm or something and it meant the trucks couldn't get to the factories on time that can be a real problem Then all of a sudden what do you think the next it's just now ending

actually what do you think the next big thing that that program focused on was processes factories yawn supply chain yeah i mean unbelievable amount of work on supply chains so these major corporations working with this program which is a joint management engineering program i've kind of said These are the things we kind of have to be excellent in to be world-class manufacturing what do you think they're doing now as yeah they're they're actually now these these management engineering students which used to have to take statistics and quality classes and lean manufacturing they still do and

and and supply chain they all now must take a class in machine learning data science and machine learning so These are the you can follow these things and industry is telling us this is what's important okay um i think i'll just skip through this but i think we can argue that there's a certainly a flow of capital through manufacturing organizations those are oil barrels there because there's a lot of resources required and all that stuff again comes out the other end with my car But the other thing here is which is very interesting and this

this would be again for another day you start with some knowledge and it increases as you go through the factory which is really something that's this kind of cool i've actually had a colleague former digital executive who um coined this term the knowledge supply chain which is kind of interesting so manufacturing Is actually a source of it's a sink for knowledge because it wants new people in but um it's also a source for it okay so uh let me just conclude with a couple slides and then i want to get to a little uh comment

on workforce but just to summarize left-hand side technology the levels of technology so when we had standardized parts and that sort of thing I think you'd agree we had flow of materials through processes and systems the the you know the gun barrels and and the stocks and other things were flowing through the you know the raw materials were flowing through the mills here and textiles were coming out the other end what about mass production same thing how about automation same thing how about cyber systems 4.0 it's it's the same thing okay so That's how we

can distinguish between technology and principles that doesn't mean the technology isn't influencing how we apply the principles that's the whole thing but these principles and um you know we sort of say why why don't you learn these why is it important to learn these i had a colleague stan gershwin who's your real authority in these manufacturing systems And always pushing people you can model what's flowing through a factory you can understand it better lean manufacturing all those things are great the difference is you can make decisions using data he used to said say not assumptions

uh he said you use data not politics it's not who can shout the lattice or who's the you know the most experienced it's This is why we should be doing these types of things and so and and again many of our students that learn these principles they're heroes when they go out to the companies because they actually have a good reason for doing things that they're doing it's the world-class standard now all these top companies are doing it that it's it's now well accepted and the penetration isn't what it should be but It should be

done and here's the key to tonight i think it's fundamental to achieving 4.0 uh and i'll get to that in a moment but even if you're not even 4.0 if you're a 2.0 operation or a 3.0 or that 0.01 that we saw in in india these principles apply and could be used to improve what they're doing so new technology without a basis in principle Lipstick on a pig okay and again it's extreme here but i've seen so many cases going back to 3.0 where companies come in and say we've got a lousy factory we need

to get better let's put in some robots or we have a lousy you know we're having problems um we need to get better let's let's do 4.0 that factory i mentioned in in kansas um their original plan As it really originally presented to us was we're going to automate like crazy and and do 4.0 and in fact they're still doing that but on a much reduced scale because after we we applied these principles they have they have excess capacity they don't do overtime anymore it's really quite remarkable okay so who should learn these everybody but

certainly engineers entering the manufacturing Positions but we kind of unfortunately have the hypothesis that most people never really got this formal training and i think it's characterized by a student we had early on in the program normally we get students that are just a couple of years out and they want to get into manufacturing but i was interviewing him and the guy had been out for 10 years and he was sort of reciting all these things the principles i said well I said why do you want to come to mit he said well i know

these things but i can't teach them to anybody i don't i don't have the basis for them i've kind of figured it out through the school of hard docs uh so he came and did the degree and now i think he started his own company he's doing quite well this is what's really important you know this is not for the big boys only it's for the small and medium Enterprises but here's one that we're finding more and more we're finding that startup companies are wanting to learn these principles because they realize that one of the

key things that kills a startup is what's called the valley of death you know which is a transition from a nice bench up top idea to something you can actually produce in volume to make money and understanding simple things like Quality and supply chain and other things really really important and um other things this is the last one i find really interesting and we're still looking at this is it important that everybody in the factory from you know from the machine operators up through the manager know these principles Absolutely good that's the answer i was

looking for um but but uh but we're not ready to do that but that's that's the interesting thing um that is not the slide i wanted let's skip that one that one's good too but this one's much better here's the point though and this really you can see this just tracks we're talking about this is a bureau of Labor data on changes in manufacturing employment over 12 years at the beginning of the of the century big decrease 25 percent decrease if you have a high school education decrease if you have high school but no college

with some college you break even with a ba degree you're doing great if you have a graduate degree You're doing really great this is growth these are not absolute numbers it'd be you know but that tells you that the demands this is the new reality in manufacturing um so what i'm really saying is that the all of these things here and the way to take care of that education is if you're want to be an expert in cloud computing related to manufacturing you study cloud computing And these if you really want to do augmented reality

to help with training and machine operation and all these kinds of things do that and these if you're working on autonomous robots this is a big one do that and do these additive manufacturing we did an additive manufacturing class a few years ago it did something really striking with it we had the students say go out make some Parts and measure how good they are and make more than one of them and you know this was like oh yeah what about that so so this is this is our this is our sort of our hypothesis

here yes go to town here be an expert in any one of these big data is another one we have we have everybody going off and now doing machine learning and and data science If they would just take time out to learn these i think they'd have a huge impact okay so um let me just have a quick conclusion um which is this and and these are some broad statements this vision that of 4.0 and it's realized in many many factories these days the factory is now this really highly interconnected connected Uh high-speed complex technical

system even even this again this small factory in kansas had a lot of technology in there all connected very complex technical system and what's unique about it are these principles the flow and the variation um and quality rate cost and flexibility okay so uh that is something that you don't ever won't forget about and that's what you Need to understand remember that the highest growth area right now is in the area of advanced degrees but we are now convinced especially with some of this experience we've had and working with some major industrial training agencies that

is what you really need to do is have the whole workforce top to bottom be aware of these granted at different levels what you teach an engineer About the principles will be different than what you teach in a machine operator a machine technician a local manager or a global manager and we must embrace digitalization completely but not put lipstick on a plague okay thank you very much [Applause] i may have worn you out but if you have questions i'd love to Address them you never said anything about actually designing before you do anything oh you're

talking you're talking about designing products yes yeah yeah um what would you like me to say about it so that's a very good point and another another revelation we had back when we started this Leadership manufacturing program is we learned the term transom engineering anybody ever heard this term yeah you design it throw it over the door you never open the door manufacturing gets it and they have to make it so another another lesson learned here is and and 4.0 does help this the communication between those dreaming up the designs and those executing them Is

much better i could have also said and i didn't put it on my list here but we we actually have said a lot about this if you're a designer and you don't appreciate these issues amazing manufacturing issues um again you're you're missing a big part of your design space here's your design constraints so we are advocating again that In fact our master of engineering in manufacturing degree was recently renamed to the master of advanced manufacturing and design degree because we realized that in fact a good part of what we did related to design and we

have design classes as well but our students who come in with a strong design interest still have to take the principles of manufacturing yeah So um um the flows of materials and processes and systems that you described as a characteristic of manufacturing i just wanted to say they're characteristic of a much wider field so the human body could be described exactly the same way flows of materials processes and a farm would be described oh well a farm is a factory but yeah Yeah so it's a little misleading to present those as being attributes of manufacturing

well i i would agree with you in the sense that that those two terms of flow and and variation apply to a lot of things i'm applying them to this thing called the manufacturing enterprise as opposed to um It's a technical enterprise as opposed to some other technical enterprise that um like like prototyping or or design is is applied to your farming example by the way a lot of the statistics and other systems things that we have here actually came from agricultural research years and years ago on the body and the earth i mean my

only thing i would say there is that having actually once done biomedical Research i went into manufacturing because i got tired of not being able to change things in the body and once i you know i couldn't make this machine i could do a prosthesis on but i couldn't make this machine but i could make that machine and i could i could vary how i do my do my uh my agriculture but very good point yeah could you touch briefly on roll to Roll in opposition to earlier approaches and to what extent you can convert

roll to roll well you know actually um i have a i have a small credential in roll to roll and i don't do it anymore because um roll the roll manufacturing you know you can imagine you can you can roll textiles you can roll paper you can roll films you can do many many things and You can make um lots of product very rapidly this way but it's it's it's again it's another technology that can be very very helpful um it's an old technology we've been doing it for for a hundred years sensors controls and

other things have greatly improved what it's doing by the way the most computerized factor i've ever seen in my life Was the alcoa factory in alcoa tennessee that makes aluminum for beer cans because to to do it at the quality cost and rate that you needed and these enormous machines were doing that so roll to roll is is is a big thing i did some work on printed electronics that's coming a long way could we do it really fast there was a startup called karnarka down in new bedford actually in lowell And and then down

an old polaroid facility in new bedford that could make solar cells that rolled up on a roll about this fast and that was pretty cool it was amazing you know why you can't buy a canark or solar cell they didn't work very well they worked pretty well but they had some problems but i don't know if that that got to your question But roll to roll is is a particularly in in the area of electronics in other active devices and other things that that is is very very important and it's it's governed a lot by

our ability to now um control things really really well for for not too much expense yeah go ahead um so what about manufacturing five points yeah i know i mean i was just thinking of things Like like zero gravity systems operating yeah only thing i can tell you about uh here i have a slide on that the only thing i can tell you about 5.0 is that where'd it go yes this will still be true [Laughter] no um we we've been we're joking because you know as i said it's it's just a progression And um

i do believe as with all of these you know why did we get to automation well there were some breakthroughs that actually had to do with stuff like developments from world war ii that led to servo mechanisms that and cnc machines or nc machines led to some basic computer programming and and you know all these things happen i do believe that that it is this evolutionary improvement in the cost and And quality and utility of information systems that's done the latest one i don't know someone brought up the star trek materializer machine you know maybe

that would be the next one i did see by the way just the other day former student of ours is at berkeley who has invented this process and the demonstration of this i don't know if you've seen this but you know a Lot of uh 3d printing is stereolithography where you shine or single beam you shine it on the surface of a polymer that that solidifies when you hit it with the light he took he takes a beaker of this same material and spins it and then using command a computer tomography he irradiates it with

the right wavelengths and it spins around for about a minute or two And i mean it's it's a miracle all of a sudden a three day three a a detailed three dimensional object appears in the middle so it's it's like volume 3d printing it's not it's it's almost not like additive manufacturing it's like look at that so things like that are happening and that's what's really exciting but i would see that more as a progression Right now i'm not a good enough prognosticator i thought we were never going to have a gigabyte of memories one

of the key elements of 2.0 was scientific management yes yes frederick taylor yeah essential principle of taylor was removed from the shop right right the essential feature of lean or the toyota production system is put the brain work That's right that's right so people who work in a chevrolet factory in framingham could never influence variation whereas someone working in a toyota they can pull the cord and say stop something's wrong that's right where is it's in the cloud i mean it's obviously in the code but yeah i mean i'm really thinking about the actual manufacturing

process no i think i I think that's the brain yeah well it's it's a good question and again we're getting at this because we have thought about how would we bring these principles to someone at the shop floor level and why would they need to know them and our hypothesis is that they could exploit the 4.0 technology better if they knew what it was being used for and why i'll give you a quick example though again i'm back to this It was a great trip to this this shop in kansas they had put in a

cobot does anybody know what a cobot is a collaborative robot it's just a robot but it's one that has has very soft movements and one that gives if it hits things and that certainly so they put one in to load and unload this machine because it was really heavy stuff to put in there And they had the guy from the robot fact vendor there and it was all great but i tell you the only people who really knew what was going on in there with machinists and they just had any really knowledgeable guys they were

up on the technology so i think actually you know it it who knows about displacing the actual um number of workers i'm not so sure about that if this is about doing it Better and and stepping back for example at my my interest which is at the process level having much better process data being able to understand what quality problems are before they come along maybe putting more controls in place to replace something that a an operator might have done by measuring and thinking uh thinking out new solutions and that sort of thing But i

think i think it's a really good point i do think that the brain is here to stay as that you know just as an aside once i i heard not that many years ago at least for both socioeconomic reasons and technical reasons there was an auto plant one of the detroit plants had a plant over in windsor canada they were hiring degreed engineers To run machines and things on the factories because they're getting ridiculously sophisticated and expensive so yeah yes sir is an electric utility a manufacturer well they certainly have variation and and rates yeah

i i i i couldn't venture a guess on that um they they are um you know we talked about converting One thing into another thing and adding value you could say yeah they convert usually a fuel into a valuable commodity called uh yeah you could call them a commodity producer yeah yeah and more and more we don't want to run them the same the whole time you know the ability to follow is really important and putting in storage is getting more important yeah supply chain yeah which we're feeling Yep yep yeah the ability to get

the resource in and out getting a good grid yep yep i didn't know my principles were this universal i can i can tell everybody now they they describe the earth that's great thank you very much for that yes sir yeah well it's interesting i i mean i i i'm going to speculate on this i have a little bit of content knowledge i Um i love to say this i drove up in my my nice new tesla model 3 and some of my students were actually on the quality team that made sure the doors closed and

things like that so i'm really psyched about it um but um i you know i don't know how relevant this is tonight because one of the things well there is a relevant thing you may Have heard the story that the delays in delivering that particular one were because i i don't know if i want to say this because we're taping this but but there was a little bit of the lipstick on the pig problem in other words they decided to blast the factory with technology beyond what was really necessary for at that time to get

the task out it Still takes time to do it they're catching up now and learning but with respect to electric technology i think my my opinion is that is still primarily a a fundamental technology issue it's it as we all know the key thing with the car is the battery tesla and others have have now you know they have their famous gigafactory they have tremendous manufacturing capabilities so they're really Doing a great job with those cells probably making them as well and as cheaply as anybody could but they still cost too much to to put

together but i think um what i what i really like about tesla i also had a chevy volt for a while is there they're really well made and the manufacturing seems to be superb and that only helps so at least you can say Well that's not the issue it's not that we're bad at making it it's that it just has an inherent cost associated with it i think you know i i that's what that's what the battery people tell me and that's what you know um all you have to do is if i suggest you

do this sometime go on the tesla website and say i want your cheapest one and then they say yeah but would you Like a hundred miles more range i say sure and they'll say okay nine thousand dollars you know well maybe it's not quite that bad anymore but yeah that that's still when you look at the difference uh in cost that's the driver that's the driver yeah but we're getting there and everybody should have one they're the best things in the world Okay great thank you all very much great questions thanks for your attention [Music]

you