Zach Kaplan on the Digital Manufacturing Revolution

my name is Zack Kaplan I'm the CEO of inventa bills online hardware store for designers I love building things originally we had a business servicing rd and design people at big companies companies like Black & Decker and Nike and then a couple years ago the cost of making stuff the cost of machinery all started dropping and then websites came out like Etsy and Kickstarter and we launched this hardware store to make that research that we were doing for big companies available to everyone the world started paying attention in a big way to digital manufacturing and desktop fabrication because the cost of entry just like what happened with desktop publishing in the 80s all of a sudden drop down to a couple hundred bucks to play it's really exciting because now anyone can be a manufacturer these low-cost digital machine do you plug them into your computer with the USB and now you don't need a fortune 500 company to make products you just need some ideas and some time and a couple hundred bucks you now have a manufacturing facility on your desktop we're seeing successful business to start up and grow from nothing with digital manufacturing as their engine so for example one of our customers makes furniture one of our customers makes wedding cake toppers one of our customers makes jewelry so they have a machine they order the supplies they make a small batch of products and then they sell them either on their website or Etsy or the local retailer like a small business retailer we're in the early innings obviously manufacturing has been around for hundreds of years but digital manufacturing is definitely in the early stages most designers still are just kind of figuring out okay so what are these tools and the tools are improving I think this is what it felt like in the very beginning of the Industrial Revolution I think this is that next revolution the digital manufacturing revolution I'm just totally obsessed with what's going on and I think that it really has the potential to change the way our economy works my goal is to build a large business and fundamentally change the world change the way people approach science and technology and change the way people approach Park development so I hope that we look back we'll see that inventa Buhl's helped ignite the digital fabrication revolution I believe that that will build a new generation of entrepreneurs who can be their own boss and get to love making stuff every day

Paul Homan, CTO Industrial, IBM UK – discusses Industry 4.0 – From Exploration to Adoption

[Applause] so yes mind pull Roman and I was very glad to hear you again refer to the northern powerhouse because they're coming over from Sheffield this morning I must just say though even though I have come over from Sheffield I am really cold up here I don't know it's light down there so what I wanted to actually talk slight change of tack actually as the CTO my role in IBM it isn't to kind of big up the possibility but actually to worry about the delivery and the sustainability of the solutions that we provide so I want to talk a little bit about what that means for me and what I think the big challenge is so hence I've titled it from exploring to adoption I want to talk about how we move from actually those sort of first tentative experiments through to to scaling 0 so just a little bit of about me as a child there were three things I was really interested in one was nature one was machines especially excavators as all kids are and explorers and and whilst I could talk about biomimicry or robots and collaborative robots to capture those sorts of things which are some of the intersections there's a little area there that I wanted kind of reference out which is the interplay between industrial machinery and explorers and there's a quite a book that's written 90 years ago I'm gonna use it as a bit of a theme and reference and some of the lessons when we look back about what we what we're looking at today and that book was written by Maurice Holland 90 years ago called industrial explorers and hopefully you'll see why I use that as a reference point there's also a small quote there from a Robert Browning poem which the part there goes a man's reach should exceed his grasp or what's a heaven for and the idea behind that particular quote and he's used with a lot of is is actually that you should be reaching slightly further than you're comfortable for in order to be able to achieve what may seem impossible today but will be tomorrow's reality so industry for what it currently exceeds in in this room our grasp what is slightly beyond what we're trying to reach and I've put a few terms up and I could have put many many more but the idea was to kind of just put some things up there and I'm sure some of these are seen as enablers some of them are seen as issues is it some of those that we see we struggle with or is it actually the totality of those the some of them if you will and in the next question is do we believe that's beyond everybody's grasp well some of the stories we're hearing clearly not some people are starting to actually be able to take advantage of some of these things and what I put up here is a very very simple value stream for manufacturing you know you think of an idea you design it fabricate to assemble it and then text and Commissioner and we've heard the term digital twin quite a few times and one of the challenges I want to put to you really is thinking about what digital twin does for us in manufacturing but actually one of the challenges which is double-sided and I've come back round to this at the end and the point I'm making is we get the idea of a digital twin a digital copy of the physical product that we make going end-to-end through the entire project lifecycle but it also allows other people to cherry-pick parts of that value stream and that's the challenge I want you to just have at the back of your mind as we go through this particular presentation so I wasn't going to major on this but you may hear this and you'll hear this in other areas I'm sure these are some of the common lessons and some of the common ideas that people will talk about when they're talking about industry for made smart or whatever in that yes you should start now you should start experimenting don't be afraid to fail you know everything there are things that everybody at all levels can do to start off the journey and you should start with small discrete projects preferably things that give you some form of data point help you with your datasets and improve their forward however the kinds of problems that I deal with and not about experiments and not to do with how do we try little things out it's actually so what once you've kind of convinced you want to do something how do you do something so that you can bet your whole business on it where you can't afford to fail so coming back to that book that I mentioned by Morris Holland the industry explorers you can take great heart from the fact that of the 19 organizations that are listed in that book from 90 years ago 78 percent of them are still in existence now contrast that with only sixty years ago only twelve percent of Fortune 500 companies listed 60 years ago it still exists so clearly you could infer that those companies are doing something right now what were they doing why were they in that book they were researchers they had lots of areas that were trying to innovate and they were trying to bring new practices and processes in the way they made things and what they made into their organizations now was it about the technologies well actually I've read what's in the book are not going to sure you know there's some rubbish ideas in there but it was the fact that actually what they were doing was having a cultural change they were willing to invest in in looking forward it most definitely but it was also something else in there which is where I come from coming from and that is and this will be no surprise to anyone that knows me that they had an underpinning architecture that allowed them to take those ideas from from that just an invention in the research lab and scale into their business now I spent 30 years so it's not many different from Lukas in this space and I have condensed those entire 30 years of clicking architectural tires trying to make things work sorting out problems that scale into just three words and this is my gift to anybody I tell every architect I ever mentor these are the three things and only the three things they need to worry about viability integrity and extensibility if you take nothing else you take those back you make sure that anyone who gives you your solutions understands that those are the three things those a recipe for success for scaling so what I mean by that viability simply will it work will it do what it needs to do okay that may sound simple but actually when you look at that from an industry for made smarter point of view two areas really jump to mind integration integration is not about plugging together technology we can do that we can we that's not an issue integration is about making sure that the data within those things that you put together play nice okay think about it like different paints if you start mixing those up you don't just want a big brown muddy puddle like used to create a kindergarten if you mix everything together it has to be compatible it has to work together so integration is absolutely vital and location and this is often forgotten and and is very very relevant in a manufacturing environment and wonderfully trivial the thing about location and we all like the idea that we couldn't sort of you know potentially know exactly where somebody is in a shop floor we could know where an autonomous vehicle is in relation to them we can worry you know all this kind of stuff – ility mapping wonderful and we think about that and a phone and you could walk outside here up and down the street and you know we're kind of easy to make that assumption however your phone assumes you're at ground level or not somewhere flying in the air it also assumes you're near somewhere you're allowed to go a pathway or a road and stuff like that and so it's approximating and of course it's outside now most manufacturing occurs inside there are large lumps of metal around large lumps of concrete sometimes it's under round it confined spaces etc and the fidelity you need to make sure that a person and a faultless truck don't cross over obviously has to be very very high so location and senses and the ability to better worry about location in your viability both is a very very key point and often often either often overlooked integrity integrity in any solution really means will it do any harm will it do any damage to the stuff that's already in there so when you put something new in can you ensure the integrity of the processes can you ensure the integrity of the data and as well as that within inch before it's also about the security of it and you open stuff up you need to know basically who's accessing it and how one of the key points though on top of that is when you do this you will create a lot of data there's a lot more than you can just normally handle so how do you actually make sure you can make use of all of that and that's part of the integrity as well to make sure that you can exploit it correctly means that you have to have layers on top that can do that kind of analytics and a third part around extensibility and I'm going to use this as a framework as we go through some examples is really around platform to something to be extensible you have to have a common platform that can not that doesn't close off doors doesn't close up avenues it can cope with future change that's true of anybody if you make any complex product you'll understand that concept and the third part and we've heard this a few times already today and it's very become a pervasive term is ecosystems now what I mean by ecosystem isn't just a random collection of people coming together and basically having to transact it's actually a deliberately organized unit that they're able to work going forward and an ecosystem has to be able to swap in and out different partners as it goes along and that's not something just happens by accident so I've got three examples and I've got a challenge I wanted to build on – yeah and kicked-off unlooked-for stories and I have to say I really wanted to put some UK stories in there it would be great if next year the stories all had to be UK only stories I think that would be a big big challenge and I actually think that's something that we should consider very cleanly in this room that my the reason for picking the three stories across a piece is another three-pronged mechanism I use around product enterprise and ecosystem I look at industrial manufacturing and basically I divide that into three things the product that we make the enterprise the place that we make it and each system is the place where the product operates and so I've got one of each of those sort of space stories just to kind of show how I think industry for affects those so first one cone a elevator and lift manufacturer had a huge install base they maintain and look after a large number of lists using the viability integrity and extensibility assessment a couple of things I just draw out viability could they make it work could could they connect up the 1 million plus lifts that they've got yes that was that was a test we had to go out and kind of do all that absolutely that was the integration element within a six month period one of the key things behind this though was around the integrity was around the data and the fact that frankly putting that data into an IBM cloud could in theory mean there iBM has got access and understand people float I'm sends a movement of people that data is connais owned so they entirely have that data and that's really really important to them and I urge you to be aware that when you have that insight that it's highly valuable understanding your information and the insights of that information that you have with it is key to these services going forward and talking to services the last part just as an example around extensibility one of the things that connais were keen to do was actually to build this developer ecosystem so other people could use their platform and there's a cab hailing firm that have built an eco built an API service so that when you come into your apartment in the US this happens to be a new presser for a lift to go down to the ground floor at the same time there's a cab hailed and waiting for you as you get outside on the ground the ground floor or street level or whatever they call it in the US so a second one and they ponies like I couldn't refer to the individuals but this is a an automotive company that was taking 150,000 photographs a day in order to look at paint scratches and blemishes and using the viability integrity and extensibility points here but key thing about the integrity that we wanted to look at was could you repurpose that same data and use it for something else in this case actually checking the doing visual inspection on a task that was performed to check whether or not certain things were performed in the right way certain tasks were completed and but by performing a testing out the viability was could we get to a greater than 95% accuracy of that visual inspection so that was tested in the small amount and then scaled and the point about the scaling is it took three and a half days to train up that capability for that use case now that's really important because that's less it's at least 14 days normally for a new use case about three and a half days per use case means it suddenly becomes viable to being able to add more and more use cases on the same data and extend it across the entire shop floor and more operations and so the last case illiterate reference was was Merce whereby there was a lot of different parties are involved Chang and I think this is quite relevant because it was about having some form of neutral access that everybody could kind of get to go and so this was they wanted to produce a a blockchain proof of concept based on an open and neutral platform that had trusted data in it from and that actually was based on this sort of idea of smart contracts that has now been produced and subject to regulatory review is a joint venture that we've gone into Merce with that will has a whole series of parties signed up and that provides that kind of element of extensibility so just to kind of recap on on the viability integrity and extensibility part you know these are the sort of terms these are the kind of the highlights of what my testing framework has used to kind of say how do you take these from ideas and actually do something with them how do you scale these and do something serious and I won't read them through the whole things all over again but just to kind of pick a couple you know connectivity is key you know that connectivity was key in an aftermarket situations that was in other words something that was a pretty existing install base and then the ability to be able to develop new services reuse existing data these things that absolutely can because obviously when most of these are not starting from a from a greenfield situation so one thing I wanted to I said I'd come back to was that point about the union's about the digital twin so this is a chart that I've taken from CB insights deliberately full at one end this deliberate what it is over the last nine years or so is it a chance of what are called unicorns and apologies if you're not familiar with that will you slightly to move essentially these are startups that have gone to a billion dollars or more in valuation and nine years ago there were you know virtually none and you can see nowadays and it only goes up until last year there's a lot and you'll recognize the air being bees and Spotify is in the uber is early on but there's many many more there and one of the things that these organizations have in common is that they have been able to find something digital that actually has challenged their individual markets so if I take an air B&B or an uber or whatever else you know we can look at the platform they've built and the digital equivalent that they have created that digital twinning concept for me is not a million miles away there is a potential unicorn sitting somewhere who can own just the digital tweening coffee and never touch metal and if you think that's a little bit too far-fetched you only have to look in the construction industry to see how close that is and so just add a wonderful timing I wish I could claim it in addition though just this week iBM has launched released its study or C suite study where we've interviewed 12 now a thousand or so c-suite executives and if anybody wants to know any more about it then you know please see me or someone on the stand and we can get some access to that for you but the key thing about it one of the key messages that type of the report is called incumbents strike back because basically it's it's recognizing the biggest thing is that people are recognizing this potential threat is real in all set and we'll come in manufacturing as well and it's about how they are gearing up and what they are doing in response to that to be out of scale of solutions to ensure they don't get left behind and become one of those 7% that I referred to earlier so thank you very much

Introduction to the Industrial Internet of Things (IIoT)

my name's Duncan McFarland I run something called the distributed information and automation lab hearing at the Institute of manufacturing in Cambridge and this what we really do is look at industrial applications of emerging automation and information technologies and this talk is is a is a reasonably elementary introduction to industrial information internet of things but also internet of things generally and I'm just going to really attempt to touch on some some basic questions and because it's lunchtime and I thought I'd take a light-hearted approach and so most of it you'll see is in terms of cartoons so you'll have to judge the artistic quality of those yourselves but best I could do so I'm going to move us on to the next slide at least I thought I was maybe I'll do a different thing but we've got a bit new technology going here thank you so the internet what I'm going to do is break this talk into two two parts first seven scanners spend the first 15 20 minutes just talking about the Internet of Things and I'm sure almost everybody on online has actually heard of the Internet of Things I'm less sure how many actually know what it is and where maybe where the idea came from and what it was intended for so just good good before before we talk about me in industrial use of Internet of Things I thought it'd be useful just to spend the first half really discussing the Internet of Things and and in some ways I've God I'm in a good position to be able to do that cuz I was involved in the very early days of the notion of Internet of Things so you'll see as we go on and I wanna spoil the excitement so then the next stick and half will be about industrial applications so Internet of Things what is it this is the most boring slide this of the webinar so it's essentially the nip nip the it's the ability to network physical devices a lot equipment vehicles any other item in and by embedding it with electronics and software and getting it in making an achievement network connectivity for those items so we can actually gather information about objects and exchange data about those objects so the thing that is important to remember because the times we lose the thing that in our thinking about IOT but it was it can't that's the origins it's about connecting things to the Internet as it sounds so yeah the first graphic there is is a simple attempt to illustrate the the basic mechanisms of the Internet of Things and I kind of gave a talk about three parts that down the very bottom we've got collections of sensors and actuators so sensors that gather information actuators that do things and with the the idea of the Internet of Things is that we we can gather information from those items or get information to those items from some form of what called gateway is the language I'll use most of the time today though there's those boxes superbly colored in green yellow and and red and those gateways will then gather data and pass that information predominately to the cloud to some sort of server location but generally we're talking about car based servers so there are three basic elements of Internet of Things so one of the things that I'm asked quite often is you know can you tell me about the origins of the Internet of Things and the you the next slide is very close to to really what the actual origin order the Internet of Things was and this was this is a picture that came from a project I was involved in Cambridge and MIT and other other universities were involved in back in the early 2000s called the auto ID Center and what you see there is a rather poorly drawn bowl of pasta talking to a microwave oven and the the this this notion of physical object we had to talk to a machine or a device was the driving point for this project the auto ID Center which was actually looking at next-generation barcode so effect using looking at using RFID as a mechanism means of connecting an object like a bowl of pasta to to the internet and using a radio frequency tag attached to the object and an identifier on that on that tag to point to where data held might be held about bowls of pasta and therefore the microwave also connected to the Internet oughta be able to interrogate that information so back in back in axioms 1999 Kevin Ashton who was seconded from Procter & Gamble to work on a project actually coined the phrase Internet of Things for the first time and the thing that came came out it came from from from that idea what which is something we worked on for the years following was was and I the the notion of what we called an intelligent product and the intelligent product is you know it was if we've got this bowl of us but spaghetti connected to the internet could we actually to a certain extent give it a way of thinking about its own progress and clearly it was very hard to put about a brain or even a computer inside a bowl of spaghetti but if you connect that to the Internet in perhaps you can put some thinking about further but on behalf of the bowl of pasta on the network so we had the win with this idea of an intelligent product which you know if the bowler pasture was had the notion of how do I get myself cooked then we identified that it needed five things I needed a notion of identity just like a person does in fact all these are like a person it needed the ability to sense its own state and and a communication mechanism to pass on that state to to the network need to be able to manage data about itself then that they're more complicated it needed a language in which it could actually not only send message but actually send meaningful messages for example to the microwave oven and ultimately support decisions so have rules and mechanisms available so this was the idea of an intelligent product and to a certain extent the vision for our vision for Internet of Things is based around that so I'm going to go to leave that vision for the moment and start to now pick up the next question which is where what's the internet bit of this and so if you look at this slide you know those of you that are old enough and I'm just about old enough will remember that internet actually began life in the US military in the 60s as a meant as a means of connecting compute disparate computers together it then evolved over the over over over the next 20 to 30 years as really as more and more computers and requiring more and more sophisticated networks in order to connect not just a couple of computers but large numbers and to the extent we where we you know in the in the nineties most most people around the world were for example using email on the easing internet then in the 1990s we we started to expand consider the expansion of internet capabilities to start to consider other devices that might be connected and tablets mobile phones PDAs personal data assistants which don't exist on their own anymore but we're very they were very excited about back back in the 1990s we're all devices that could be given network connections in the same way that computers could have a network connection and then moving on and you know these data are rough in fact almost everything in this talk is rough but you get the idea you know in the 2000s we became more and more interested in how we could get machines and equipment and you if you're struggling with the the images that's a car a robot and the fridge there but so devices that were powered but ordinarily not necessarily connected to computer networks so we looked at me and isms whereby weather by embedding in computer capabilities onto those devices or at least in having reasonably sophisticated communications we could also give those items IP addresses and have them as part of the internet family if you like and then this latest stage is is the where we've moved to things so whether it be a car tire a chair a suitcase or a piece of cargo mechanisms for getting objects everyday objects which mail may or may not have any powering computer or computer capability but by the use of technologies such as radio frequency identification wireless sensor nodes Bluetooth tagging etc we can come up with mechanisms for connecting also those inanimate objects to the network and that's really where the notion of the Internet of Things comes from it's and it's more or less an extension of the Internet to the to to expand to taking to it everyday objects and the one that one comment to make there is that obviously none of those objects ever themselves connected to the Internet so they normally are going and putting through a third-party device such as an rfid reader you see on the right-hand side that we point to that here rfid reader or through a Bluetooth reader etc for example like a car tire talking through the local communication system on a can on a vehicle ok so it's that's the backdrop the backdrop to you know where the Internet of Things and the Internet can bar come together they they really are part of the same thing the Internet of Things is ultimately just another step in the extension of the internet so to reasonably technical it's just technically as we get today to to technical issues to firstly what what is the Internet and what it was sorry what is the Internet of Things and what isn't the Internet things so first put first down the left-hand side in blue you know the these are these are the points I would make and maybe worth remembering in terms of what in there of things is it is an enabling infrastructure for connecting objects and things and entities to the network it's also there for because we can connect them to the internet we can therefore connect them to each other using the Internet as a medium for doing that I it is an extension to the Internet and it enables sensing which is where a lot of focus has been so far but also decisions and actions to be made involving everyday objects by contrast the right-hand side then says what what what Internet of Things isn't it's it's not an it's not a new application or a new service although it might be the enabling infrastructure for new applications and services importantly it's an as it stands for today it's not a new separate network of objects the objects are actually being joined to an existing network and that's that's important because often we don't just want to connect an object to another octave you want to connect it to a machine on a computer or a device so it's actually massively beneficial that it's part of the same network it's also not just about getting sensed information from objects actually there's very limited value and benefit in just gathering sense data on its own what we really need to be asked there is to analyze data and make decisions about that object or other or other entities and then put in place actions that benefit from that from that informed decision equally it's not about big data or it's not just about data analysis there certainly there's a potential forcing for significant data analysis to be performed but often these two things are used interchangeably and it clearly you know there's quite different emphases and it's an illness not also about database coordination you know where having said that we require database and data management but as we'll come to in a moment that's just part of the picture so an Internet of Things is not a technology as such either it just enables a lot of is enabling lot lots of new technologies and one way to think about the types of technologies it's enabling is the next couple of slides so when when we so entitled this what happens on the ground and what I mean by that is if you look on the left-hand side of the slide we've got information coming off an object it's being sensed transmitted received at a base station or a gateway and then being trained to put onto a network and then and that's the sensing part and the the actuation part or the acting part is from the network transmitted and through a gateway transmitting information to a node and whether it's received and an actions taken so how do we well there are a lot of mechanisms a lot of ways in which that type of process can take place not just one approach and there's a whole range and this rather hastily drawn sketch tries to pick up some of the different mechanisms that are being used and this in the wireless space and I should emphasize into the things it's not just about Wireless either but a lot of the developments you hear about are wireless and and so you know the developments ranged from let's go to the far right hand side from no power low there's one called low powered wide area networks where since data has been gathered over up to 15 kilometers and being transmitted infrequently and very low powers to a base gateway to or the other extreme of that the low bandwidth in the body area networks where people developing ways in which you can have a a little here at a network based around the clothing you're wearing or the sensors embedded in your body I hope you able to don't have too many of those at the moment and then at the other extreme and perhaps into the space where a lot more of the industrial stuff operates we move up through RFID – where you're getting more bandwidth and therefore being out should move better greater data volumes at data rates and to you know use of Wi-Fi and be the evolving 2g 3G 4G telecommunications communication mechanisms so range and and data volumes is just one way of thinking about looking at this but but it's important to look at it because for example the ZigBee which is a lot of bit for example a lot of home developed home wireless sensor networks have been developed around is just as much as Internet of Things development as RFID Wi-Fi or low powered wide area network so no one of them has an exclusive exclusive hold on IOT so that's what happens in on the ground and I guess the next bit that goes on is what happens in the cloud so once that data has been gathered there are numerous organizations providing cloud support for for that data and again you know we hear a lot about data analytics going on with with data that's been gathered from sensed information associated with objects but if the reality is that the jigsaw is quite a bit bigger than that and I just picked up six of the pieces here but you know from a very you know for example we need to be able to link the data that's gathered with the object that the data's come from and perhaps other other similar objects and from for that we need to be able to catalog classes of objects and be able to search for them and where they're being used going down then we need to be able to gather queries so because if there's going to be if these this data collection is going to be useful then should be able to respond to queries being made and we do need we are looking to analyze data and discover trends predict future behaviors etc and then down the bottom reporting is clearly critical but equally critical on the bottom right hand corner actually be how to gather rules associated with the informations being gathered and there executing controlling controlling commands that would go back down to actual actuators that are connected to the sensitive information so this and more are the sorts of things you'd expect to happen in the cloud associated with an Internet of Things infrastructure so when when is IOT useful well that's an interesting question and it's it's it's some you know another question I was asking a couple of years ago was is IOT useful at all and and having having looked into it quite a lot more since I think you know there you can start to categorize situations when just your standard sense sensing infrastructure that may be in place in a building or in a home or in a factory may not be sufficient for the type of needs they're required and so down down the left-hand side there I picked out just some of the categories which might be boxes you'd want to tick if you were thinking about should I make an investment in getting involved in Internet of Things so if there's a if there's an application that requires data to be gathered from multiple organizations then sometimes actually having a mechanism which gathers information from beyond the business information system environment of an individual organizations is very attractive so it doesn't get you over the security challenges but it maybe unifies the security challenges for that application secondly if if the system is naturally distributed then that that can certainly lead to some advantages in considering a system of a sort we're talking about the little diagram that in a picture on on the right there just illustrates some some work that we've been doing looking at distributed management of devices and objects in the home and and and connecting that up through an IOT environments to maintenance services but and also it but also allowing individuals that work in the home or use the home or visit the home to make comments and add information to that asset management environment also when when the software is perhaps naturally cloud-based then then there's a there's a neat link to making you submit new things so I think they're the three main characteristics but I put a fourth on down the bottom and this is something we'll talk a little bit about in the industrial piece is is where the application involves you know some degree of customization or personalization of the of the of a product or an order then actually how to really accurately manage the information sensed information about that object might be important so just going to move a couple where is IOT being used at the moment well I think that the area we can get that gets a lot of publicity is around cities and clearly cities are a good candidate because they tick all the boxes on that previous slide so the little picture there just talks about some of some of the ways in which you know IITs being used to manage traffic intensity logistics railway systems etc so just you know there and even even traffic lights and so cities are net cities naturally don't necessarily have a centralized information management infrastructure so it distributed information the infrastructure like IOT is very helpful in fact could be even more helpful if we could add some smart AI based feedback to go with it which you know if I have to take that I've got some problems with traffic players being able to drive advise drivers to take alternative routes or if I've got and logistics challenge being able to find an alternative logistics mode that sort of thing okay so just to summarize the current IT developments before we talk about industrial applications yeah there's a there's a lot of effort going into what I call the low-power long distance so the ill pwan item I had on the earlier diagram because an attempt to build some infrastructure that can be can get a kind of a network coverage for things in place but there's also quite a lot of work going on in the background in on the cloud-based issues for example around cataloging in search mechanisms and they're equally important they probably don't quite get the publicity they deserve and and the last point is that right now there's quite specific stuff going on in some focus domains so a lot of lot of work on smart homes in smart cities and we're just starting to see industrial IOT get get to get start to get some attention as well which is interesting because actually we've been sensing things industrially for a long time but that may be part of the challenge as well as the opportunity so I'll move now into the second part of the talk which is industrial IOT so you know again I'm going to just cut across the top of some of these questions the obvious questions you might ask about industrial IOT and you know I think if you find these these this I've started to answer the question but not completely answer and you probably right and have you take some questions at the end but also just to mention we be running a full workshop on this issue in the new year so what is industrial IOT well let's let's let's do it really simply we've already got a definition of IOT so it's just the use of Internet of Things to create value this is important because industrially nothing really gets off the ground and this it creates value but to create very forward not just processes industrial processes and their supply chains but also for the products and services that are sold from the industrial environments as well and and there's an attractive aspect there because IOT gives you this mechanism not just to put sensors on processes but on objects and products flowing through the supply chain so a couple of couple of diagrams to maybe set the scene this one doesn't look remarkably different to the slide we looked at before but in a sense what I've done here has just literally added some industrial objects are down the bottom and and also noted that actually you know although we talked about Claire there's nothing that precludes cloud being cloud is really just another way of describing a server that's like the server capability that's located across the internet but server capabilities locally can actually perform some of the tasks that we need to do anyway so that's that really in it's a minor extension on on what I was talking about earlier on and I just wanna spend any time talking about the way Manufacturing's involved I just put in this water well sorry I've forgotten I had an extra slide here I've just put this slide actually just picks out a vision that we developed about five years ago from what industrial IT might look at so on the left hand side we've got this this called that if you can read it it's quite small it's called the day and a day in the life of a blender so on the left hand side we have a blender being made and having conversations with various machines in the manufacturing environment about how it's going to get now moving on to the right it gets goes through logistics chain works at how to get itself delivered to a retail outlet where it's bought and sold on the right hand side top and and maybe the retailer based on that sale will trigger a reorder of the blender then in the home lots of excellent smoothie recipes no I didn't I couldn't draw kale but kale is my daughter's favorite smoothie item for reasons I can't explain but anyway you know get downloading smoothie recipes from the internet potentially triggering repair processes when the motor burns out probably from chopping too much kale and and then and then potentially in the recycling phase actually the blender being out of trigger its own recycling processes so that's kind of our vision for industrial IOT and as I said I wasn't going to say too much about manufacturing but there's another one piece that's worth bearing in mind when we when we start to talk about industrial IOT is that the the image we've had traditionally of a factory being in a single or in a single organization the single building single location is changing rapidly and actually the the business of turning specifications and more materials into the final parts and and products and the services that go with it is often now far more distributed and when their distributor it means there's a greater independent inter dependency than there ever has been between multiple organizations more or less tasked with doing the same same manufacturing operation so I'm just out what I'm going to now talk about is when could industrial IOT be useful and by the way I would say thank you to those people that are bringing typing in questions as we go and we'll certainly be able to leave them till the end we'll try to try to pick up some of them at the end of the seminar so I just want to talk a bit about try to motivate when the industrial IOT could be useful so you know if you leave this slide looks rather blank but what if it's going to build up so don't don't be disturbed but you know what won't being pretty good at today in today's manufacturing operate operations without IOT of any sort is is gathering operational data we know you could at you could argue that we don't gather it's insufficient volumes physician for sufficient periods of time but in terms of influencing performance that there is quite a good track track record for making use of operational data and perhaps I've given an I guess I give that a green whereas I give it a yellow and amber to the notion of use using operational data for improving our systems so I am I'm getting somewhere with this don't worry so if I then expanded there and and I to include auxilary data around the operation so this might be information about our working process quality data operator gathered data or machine condition monitoring data then again you know I think we make some use of that in system form it's well we don't really make great use of is that information in operational operational performance so there's an opportunity to do better there and then I expand further to the broader factory context again there's lots of data available in in most manufacturing operations around energy of energy management stores data we've got we've got a strange pictures but hopefully hopefully you guys don't see me flashing I'm flashing on my own screen anyway I'm gonna keep going day to day different people operator data fish good stock availability what the state of tools are being used in manufacturing environment are where where the raw materials are so that sort of information which critically affects operations is normally you looked at with from a scheduling context and planning context but actually could really could affect or hey hey you're operating on a daily basis and then going further out into the supply chain you know there is information out in the supply chain from for example from the transport it's transportation systems from the demand of supplier demands for goods arriving etc and again we we make very little use of there in terms of our factory operations which again we kind of assume our being decoupled from that information and there is some information used in improving the supply chain systems and then even going further again you know the the the what I call the ecosystem information around around the supply chain you know the weather information financial data about prices traffic information environmental information some of which has the potential to affect the operations and some of that could be affected by operations but at the moment I would give them both a better read both in terms of its impact on the performance and also ability to improve their supply chain so what I've tried to do is build up a set of data environment that aren't necessarily available to today's factory operator in terms of improve there are their factory operations and and therefore you know if we could find a mechanism to allow us to readily introduce those additional data sources there'd be some significant value industrial effort in doing that so you know and III saw one exxon asking the question earlier on you know when is it that the industrial IOT can be useful well you know i haven't got the specific instances but these four pointers are at least some it provides some headway answering that question so when you know for example when when there's value in integrating data from our suppliers and logistics providers and customers because we know it can affect the way we operate the day-to-day basis that's one potential benefit secondly if new technology or peripherals are being introduced into the manufacturing environment or new tools and equipment on a reasonably regular basis and the company is quite complex to integrate that within existing IT environments that's another another but a better a better for the area or as i said earlier if if the production is naturally distributed so the data from one side actually could be readily used in the operational data from the other side oh all whether we had finally where information associated not with the equipment and the factory and the operations itself but the products and the paths we information associated with those items so let's talk about the the temperature of incoming steel billets into a into a steel rolling mill they massively affects the productivity of the rolling mill if we knew that data in real time were able to adjust a little but we could make it take that to that so the diagram on the lit on the right picks up it has two sets of arrows there you know it what one one the left hand arrow says is saying really is there information from our environment from the ecosystem nor the supply chain or just the area around the factory operations that could impact on the on the operations itself and then the center sort of the reverse question is to what extent could the operation of the overall supply chain be affected not just by the aggregate performance of the business data but actually specific operational variations they get a current daily day basis within the manufacturing environment so that kind of is that that's kind of the you know tried that tries to sum up with sort of the two-way street in terms of where I 80 could can provide benefits not just in that central circle on its own but seeing that central circular circle as part of a broader system one in which we really decouple it from generally today so the the the last question I'm going to talk about is how we go about how we might think about integrating IOT with existing within existing industrial IT systems and you know I guess you know the real the point in that first slide and you know apologies if this doesn't exactly look like your industrial IOT architecture but most manufacturing organizations Industrial Organizations had pretty good systems in place for managing their industrial IT systems so you know in this what you're looking at there is a typical well semi hierarchical structure which gathers information from sensors and actuators like switches passes that through to it might be connected to two programmable logic controllers or two robot controllers drive controllers convey controllers reactor controllers etc and in the end them and the systems they then run on PCs or servers to manage the overall operations of the plant now I realize that's a super simplistic and I actually have to give a student lecture to Marilla where I give that all in a much more detail but for today I hope you'll forgive me for giving you a very simple view of existing industrial IT so so what we've got that existing IT we don't want to throw that away so how do we think about adding Internet of Things within that environment well first point to make is that actually you know there are although wireless isn't as heavily used industrial as human industrial you might imagine because as issues with interference and potential safety challenges there there are capabilities have been around for quite a while to provide wireless gateways so the devices such as mobile devices or PCs could be actually connected in wirelessly to industrial and IT environments so they exist already but then if you'd like that's a first step towards connecting some internet-of-things objects and devices secondly incorrect increasingly organizations anyway regardless of whether the embracing Internet of Things are starting to use cloud both for performing analytics of the business data they have ever access to already or actually in some cases and there's a bit more experimental some aspects of control where distributed operations are in place it's actually being done at a high level in the cloud as well so cloud to me that's an next piece in the jigsaw then then a next step would be to consider what there are IRT gateways that are being developed specifically for free for use within industrial environments and their specific focus has been about has been in integrating for example and these are very bad pictures but different tools that may be different tools or storage vessels that might be used in the environment so the there in a way what we can then do is actually take data from those tool could recondition monitoring information competing usage information and integrate that into that existing IT environment really in a relatively straightforward manner and equally we there are other devices or similar devices that enable data from pizza piece of equipment that are being used in operations so the robots or the drives to gather additional information about them vibration information energy usage information and actually they can go on to the local Ethernet the network within the within the within the factory and so you can actually just add a little piece of in IOT capability to your system as a small step and then moving on the big picture so as well and probably the because you could get bigger but my graphic skills can't get do any better than this but if we now take that Factory that we're in put in the small box in the center you know what I could then start to do is to make use of some of those low-power wide by wide area network capabilities to gather information from sensors that might be dotted around my supply chain or even the traffic data in the vicinity of my factory and and also make use of some of the cloud services that go with those some of the proprietary some of them are open access so that's on the left hand side and then on the right hand side are potentially making use of I'm home area network so IOT capabilities in the home and that's meant to be our blender turning up again there as well as a vacuum cleaner that might be a might be providing information back to the manufacturer about faults that have been detected or about a potential improvement improvements that might be made so your usage data so what you can see is where what we didn't it of things picture is not quite as clean as the one I started with with the cloud hanging over a whole lot of sensors and actuators in in the industrial environment you know we've got a lot of sensing already we an analyst and a lot of it's wireless and actually a lot more that's wired and all of its decent sensed information and what some of the challenges are around getting the sensed information on the network actually is in some quite a lot of operations one of the challenges is getting data off our or industrial operational systems into our business systems and again there's mechanisms being provided which can support there so that's kind of the the picture for integrating in net of things and industry and I realize that's not the final answer but anyway that's where where I'm getting to today so just to sum things up you know it's clear from the way I've described that neither the IOT more industrialized here amateur field and yeah actually only have to look at the projected growth which sometimes I don't know where the data comes from for that but that's a different point but it's it's always increasing rapidly secondly there's the you know industrially in the industrial domain particularly and I haven't talked about these at all there are security challenges and privacy and data IP questions to be answered and and actually there are good solutions and so not complete but good solutions becoming in place and also being able to understand the benefits is challenging and hopefully at least stimulate some ideas about thirdly there's real value of being able you know the potential value could be split into two bids well you know one is being able to integrate new data streams to add values manufacturing operations and the other yeah if you like it which is kind of it's an honor ot thing but a be able to account better be better gathering existing sense data more easily and more flexibly and potentially making use of it in different locations so in some the bottom line for industrial a IOT is that to me it extends existing industrial IT networks which I didn't mention there clearly in existing industrial I I t8 networks already embraced internet capabilities as well so it's not it's not contradicting the earlier comment about higher IOT extending the internet and you know there there's more questions we could deal with as I mentioned earlier we have a workshop in the new year which will tell you about soon and potential some potential future webinars on on specific issues so an example I was being asked this week about how IAT and blockchain work together and that's certainly for another discussion but we you keep watch the space we may have more to say on that so just to finish back to our blender you know and and back to where I think we stand in terms of Internet of Things and and particularly industrially things in terms of that big initial vision so for blent you know that in turn that intelligent product the 5 stage intelligent product definition I would say we've got really ticked to two of the four boxes five boxes so far we we've we're capable of gathering identity of objects and things getting better and better at sensing and communicating the data that's gathered the data management is coming along significantly now a lot of cloud-based work ah a language which allow that blender to talk to the robot that it's assembling it is for example not not we don't have any definitive work there and a lot of standards work needs to be done there but actually you're the interesting thing maybe IOT and AI as I put it in the C slide earlier on how can we best support decisions that need to be made about knob checked and that sounds a bit Star Wars like but actually in some cases that we look at the only the only way for an object to get things done till it is for it to look after it itself they're they're very good examples for example in infrastructure assets where there's no other cost-effective way to do it so I'll call it into the the seminar at that point but we've got five or ten minutes to address some questions if people would like to pass them through okay well while I'm waiting for I've sort of seen the comment about blockchain which I think that someone said it's the most inefficient technology it's ever been developed yeah I think I agree with you but I'll have to wait till the next webinar definitely so I've just picked off some of the questions one of the most common senses I've seen for IIT in order my automobile in aerospace industries well you know actually you know it's I think what I would say it's the sensors in those industries actually really well developed so they're all you know the sort of sensors for that that managed most of the machining operations assembly operations are already impaired and you know you could say they're IOT because they're part of a broader family but I think that the most common sense that I see that's being added as an IOT additive is is typically the systems around canvassing condition monitoring so pressure sensors vibration sensors temperature sensors are associated with a piece of equipment which might provide an early warning to a loss of production capability is kind of the typical sensor applications I've seen there one of the main challenges these sensors have in capturing the data well I think it depends on with if we're talking about the industrial sensors in those industries the you know one of the main challenges is is getting working and consistently working sensors right in place but also the point that the challenge is actually getting the information from the operating environment where lots of data is gathered but also not passed on and getting them into the cloud and there are as I said earlier there are some solutions to the addressing the gap between the data that's gathered operationally and and actually make be able to make use of a business basis is there any particular space within within industrial space that IOT is more used and starting to get benefits as candy poles candy I think a really good question I think the jury's still out but I one area that seems to be incredibly well-suited to industrial IOT is the oil and gas industry because they're you know they tick all the boxes ansan on the criteria we talked about earlier you massively distributed a huge pipelines refineries full of miles kilometers to hundreds of kilometers of pipelines and systems and the native data gathered data each any one of those incident senses about though that equipment might affect production which is very expensive so I thought if you're asked to name an industry I would probably say or oil and gas house with Industrial IRT systems be designed as Dennis Sayer so I write my comment would be Dennis you know incrementally so start with the systems you've got look at the capabilities look at the sensors you need and and where which additional sensors could provide benefit and then actually don't don't look at a IOT solution until you know from the other point of view of your organization what's the most straightforward way to integrate that sensor you know it may be just possible to integrate it through the existing programmable logic controller at work you've got a place for example or your your energy you may have an energy management system in place that will help you to use there and but it may also be beneficial to consider some sort of IOT gateway it added to your environment which could perhaps just connect straight into your IT systems infrastructure so okay what are some of the both what are some of the most effective mechanisms to understand the best industrial practices in IOT implementation really good question I I don't think I've got an exact answer to that I've spent quite a bit of time talking to some of the systems providers in this space and I probably I would say the automation system providers rather than the bespoke IOT providers a number of them have actually on their websites good case studies there they have to be broad they can't get a lot of specifics away but it will give you at least an idea and I won't go through naming that the automation services terms or service providers but you can probably guess who they are okay a couple more questions what is the difference between industrial I ate in industry 4.0 okay well that that would be enough that's another seminar and to my mind in industry 4.0 if I sum it up in a nut is about three has three different dimensions it's about addressing issues of real time information and optimization in these in in vertical integration of the business information system in horizontal integration between organizations and across the concept product life value chain and so to my mind Industrial IRT has particular advantages in in the second in that is providing horan really effective real time information integration across on a horizontal basis so between organizations and their suppliers customers and other partners in their ecosystem I think most people would say it's a contra contributor to industry for not a separate a not an alternative although I know there's a lot of work going on to blur those definitions so you need to look rather carefully which areas of industrial IOT require more research than others that's that's a interesting question I think in a way that last slide which is still up there maybe says that and I'm not saying that the the deals done with items 1 & 2 in fact actually item 1 & 2 together are interesting because they're quite often we get sensed information object identity information and we can't connect the two together but but I would say on average it's the it's the area of taking information down back to objects and devices based on information has been gathered is where we might want to focus our efforts so around you know coming up with genuine ways in which objects and devices can communicate with each other a sort of a Babel Fish if you like for object conversations and and and how had the most effectively and cheaply and logically support decisions that need to be made for an object particularly when there's you know billions and billions of objects we don't want to have them all having arguments with each other so I think there's some quite a lot of challenges there is real-time information also valuable the answer is are not sure quite where this is Julian one who's in Korea hi Julian so real time for yeah I mean real time information is definitely valuable and I you know I think you know I would sort of split you know there's real time time critical and non real-time and and time create real time is you know sort of dictated by a clock speed and that that can be important I think most IOT applications today and not looking at real time but some are time critical in that they're try their data is being gathered in time to be able to make a decision which might influence the evolution of a set of operations so I think real time information is valuable it was probably less work in that space perhaps one more question what what is your view on the emerging commercial IOT operating systems from the big vendors right there that's that I wish that was not a simple question to finish with was it so I mean there I think it's I think perhaps the previous slide signalized slide which said it's an image to your space sums it up I think no one organization has a comprehensive IOT offering for all aspects of variety as we've discussed and today I think the thing to be looking for if you're if you're trying to choose between them is to what extent you know hey you only the data that gathered in the systems and secondly to work to what extent the the system is open and accessible from a third party users because actually a lot of this is about is is about permitted information sharing and that would be the other comment to make is you know though those providers which provide a secure private access for data it was certainly going to have an advantage in terms of the industrial space I think that's probably time to call to a close so just final comment is that you know I don't have been quite a few other questions sent in and thank you for all of them and apologies for not answering all what we would do is look at the other questions that come in and possibly broadcasts on the on the website answers to some of the other questions that become available and also just stay tuned for some further webinars and a workshop in this space into the new year great then enjoy talking to you all goodbye

MxD Future Factory: Discrete Manufacturing Testbed

so we are mxd the nation's digital manufacturing Institute we focus on cutting-edge research and development in digital manufacturing design we exist to help innovative manufacturers forge their futures and ultimately help strengthen the competitiveness of the US manufacturing base so one of our driving initiatives has always been in the development and integration of digital technology that has actual business value and there's no better way to do this then utilize our 22,000 square foot manufacturing testbed facility in Chicago Illinois so at this facility we have set up multiple manufacturing lines where we can demonstrate the risk and validate digital technology that comes from our R&D portfolio and the commercial marketplace since 2018 our discrete manufacturing testbed has been operational in which we're producing these mxd tokens that visitors can actually walk away with so at its foundation what's happening here is we have a CNC turning station that's producing these tokens and an autonomous robot grabs those tokens and transports them to a variety of post machining stations but that's not the differentiator here what's very unique about what we do is the data streams that we're collecting processing and analyzing this entire cell is overlaid with a variety of sensors that are going to fuel different analytics packages so that we can showcase the value of the digital threat through use cases like predictive maintenance digital serialization and traceability through that digital series ation we look at the unique microscopic surface roughness of each token and we use that to create an individual fingerprint for that token it's a unique identifier that we can then combine all the production metadata that was collected and then look back at the token and be able to see what was the temperature of the factory how will they do the machine actually perform the operation how long did it take to washing dry so there's a variety of options to collect that you know but then you have to aggregate it you have to bring it together into into one source and then it's all about what do you do when you have that data how do you process it in a way in which you're going to be able to drive action on the floor without driving action there's no ROI so it's all about how you can do those steps because that's what digital is about so we're looking at ways to continue layering on projects and adapt this platform so that it doesn't become a thing of the past we wanted to continue to offer value and to do that we need to constantly communicate with our manufacturers and industry service providers and academic institutions around what problems we need to tackle next what are the problems that are too big that they need to look towards an institute like mxd to help solve you