Tomorrow is too late! – industrial control with real-time data analytics – Neil Rothwell

and I should start by saying that I really shouldn’t be here I’m corrosion engineer by background to find myself in the 3gs the chances to try to get somewhere I never ever thought I’d be cuz I clean the other side of the fence for a long time I’ll get lunch during this this talk a bit about down ground I’m across er engineer cycling oil and gas worked in a lot of asset rich and that’s it intensive industries and I am also pretty sad because I love data so for the juvies live here I work for a company called c3 global which was directly 18 years ago and it was acquired in december this year by so fast you by bentley so what I’m going to talk about there’s a whole new sector that really has been defined by Gartner would use these sort of area operational industrial analytics you know it doesn’t really fit with something business intelligence he doesn’t fit with the other analytical sectors it’s certainly not advanced analytics it’s about using data in the operational environments of making sense of it so I’m really talking about that I’m gonna saying why now because it is interesting just how fast this this market is he’s growing and being defined row shelter in garnered one of the leading and analysts and fellows and he reckons that in his 25 years in Ghana this is the fastest moving sector that he’s ever seen he’s growing at showing it so quick that he’s struggling to to sort of keep it terms with it this would market well yes briefly some examples then what are the barriers to adoption because they’re always barriers to adoption and also is a remarkably advanced analytics the modeling etc etc I’ll deal with them what I’m not going to deal with is any of these because corrosion engineer opens up about big data so I’m not going to it’s an interesting that if you look at our industry section on the operational side of the world without doing the cantata you know it really is a blessing say so he’s talking about you nobody’s doing you know it’s it’s just not happening in the industrial sector it wouldn’t do but you know I’ll give you an example which welcomed onto later we have refrigeration monitoring system which is the loose in the world so currently is about 600,000 hypermarkets the in this year would we have fallen out to five thousand five markets at the moment collecting data from every are several refrigerators the age of fact that licensing anything to do with it which we 225 million native resolution and we’ve got ten years worth of data four terabytes you know it’s not big data you know it’s its operational tactics a different market those hidden ebay assistant manager you know we’re talking about collecting operational data so these things here are hobbits offer I making mention of just in case on display finger so one of the areas that think aunt who picked up around and garnered certainly the first on this is sort of all area i t OT convergence you know it is difference and it comes from completely different sides of the fence you know it’s– all about delivering business information it’s about managing our systems OT is all about the operational side of the business and we know we come from different areas quite often operations you go through a project design you design a control system it’s delivered it’s got touched again so you get what you’re given so the data session indicate what’s the daily astorian so i’m going to change you given and really there’s been a big big sort of barrier to to crossing this this chasm that I Quixote and I don’t think the comfortable bed fellows yeah I’ll come to this the next slide I said the slope there’s a point two years ago when I was called into to conocophillips in Aberdeen and I thought I was gonna get a dressing down there’s a sort of days we picking up we probably should have been picking up and we can have back up with engineers to yes I but if it wasn’t actually that I was told by their cio that he got a new remitting life last ten years he’d been tasked with getting emails out on time you’ve been tasked with making sure the networks data and information float around the organization and that meant that he’d actually told anybody know so before the team took care to ask him he said no no he can’t on that day to know we can’t do that he says it’s a whole new world being told to say yes so actually mean this old to go and find people to say yes to but she said oh you can’t get much more of a change than that so he was actually inviting to say write what we need to do about this and the thing is that going to do this without talking it up this is the you know what I consider the IT side of the world you know it’s all about the year the updates the CRM systems about managing that takes an essay p systems all either so listen my weed control its control its information its machine insist things have turn on and off it spawns it’s you know the information that comes out

the answers but completely different now if you’re going to link them you’re going to get a lot for lot of data clarity no it’s about putting data into context visualizing it making decisions on it and only then you really get into the sort of the whole market of prediction of descriptive now once you can sing that takes or watch you can mix all this data with all this data then you can start to make sense of it and I think the government quotes quite an interesting what there are few companies came to providing application says that you know both statement this is this year and they’re making these letters it exists in all industries we’re all fighting the same problem trying to make sense of the faceless out there and you know what’s being done well we’ve done all the yankees we’ve done the CRX you know there’s been an awful lot of implementations I don’t think I’m any 80 100 companies that haven’t gone ASAP system so they’ve all been in presented you know we need to make industrial savings now the next area of savings is the operations and we need to address that there’s a company coming I’ll deal with some of the their own gas stuff but that’s a shocking number when you think that the efficiency and the availability of Wells and the Dulce is down the sixty percent I know that’s that’s incredible sorry industries well as across to talk about 60 762 excellent other one to get back to active service greatest dirty and I’ll deal with that children but I mean when you think about that number there it’s not difficult to see why there’s a lot to achieve and a lot to be going out and I was in cold last week and we will look at manufacturing systems in a bit of a public the operators of the plant is about a pound of classical the operators sat down at their own virtual not time time management the Chairman he had a vision the operators reckon them by operating is fat more efficiently unexpectedly out by the way they can still get another thirty percent availability text system now that’s the operators who say give us the tools and we’ll deliver it but if vessels are dirty we don’t know about seeing the couch out kind of got the dealer stop we can’t charge a grandmother aggressively in the right time and we can’t know the recipe so it’s all about you know just getting information it’s not clever honestly it’s nothing we ever do in operational analytics is clever it’s just about doing the the stuff you need to do the right time it’s common across all industries I’ll give you a number of different examples tonight different industry sectors oil and gas pressure care refrigerators one of my favorites is semi live citic I know far too much about refrigeration there are 34 failure modes on freezer and people 28 desperate software that’s the example it is so you know that which is what about supermarkets now just looking at what’s happening look at the fabulous I also driving on the roads were looking at the other head tell us a natural grip bogeys is really works at this is a valid question it’s changing you know CIOs are changing the whole attitude to Texas changing and if you look at the sort of gather those which is all polished and see all the references tab here you get a slice then the growth in this sector it’s phenomenal now the important thing is that they’ll in 50 cent growth per year going forward to 2020 so 20 billion spent on exposure over to 1.3 1300 billion in what is about five years time so this is a massive area for growth and it’s not just one industry it’s across the board it’s about understanding how we’re going to drive operational analytics to deliver savings to companies and be more efficient the interesting point about this is that gartner reckon it’s new money so that’s good news now this isn’t shifting of other IT budgets this is new money which is investing systems to deliver improvements in operations so that’s great news of all of us there’s nothing lots of money around and we should all be able to get some of this so that’s great what are the problems the problems in historical really we’ve built different systems we’ve built control systems of think dealing time series data you know there’s a negative black oxidation from the operational point of view we have structure detected which cold what we need to deal with in instructions to her my you know we need to take it apart and put it back together again so you know dealing with just an asset register in sa p that might not be the way you want to see from engineering projects you need to face and take the party and Link different systems and traditional systems aren’t ideal for analytics you know they’re robust some of these guys systems that are running 15 years they don’t go down they don’t fall over but you can’t really change that but not designs of you chose for their for operations and the other thing is that the text be a lot of domain expertise so you know the amount some really clever stuff in pharmaceuticals but it’s that

transferable across to other industries well yes it is but typically lives of these vertical niches on the moment so a bit of you have a date then this is really insane this is about sorry it’s a but if you look at what they’re analyzing you say we’ve got normal time well that one I got all of these sectors across the top here government education retail banking health trusted one thing that’s the rest then you look at what the collecting down here so even back in 2014 you know we’re not really looking at operational we’re not looking at oil and gas when i’m looking at power generation of power transmission be finding water jetting up any of these utilities they’re not even part of this so all this analysis has been really into a dictator says transactional systems not really where the operations guys up but this is where we are now so this is one of my favorite slice this dates back set 1929 and somebody slaughtered on a piece of paper the Bulls I told you don’t think that’s on the water temperature and the birth rate of pipes and this is actually the Greater London area and love and behold there’s a correlation between the failure rates on pipes and the temperature way rent freezes they rupture and they leave that’s 1929 the first water company in the UK put a system in place which looked at weather patterns and determine what workloads across all of them different sectors would be was last year so it’s taken 85 years to basically put a system in place which up kicks a bit more clever than that but we’ll look at the weather will look at historical data and will come out with well because we’re going to see this trendy whether we’re going to need more people in road repairs were going to be more people than this working it ball call out in that area it’s all going to have an example in committee it always happens so X 5 years but we knew about it 85 years ago yeah why don’t we take that long it took that long really because there was danger all over the place so this is an operational tool which runs in real time and allows them to model based on historical records to the last 10 years what they can expect in terms of call outs in terms of failures in terms of all the different disciplines for maintenance requirements across all of their sectors across the whole territory now to do that I think is about 40 different interfaces and ed putting in place so they had to put one big operational data store in place so it’s an IT project so it’s a big etl system you’ve got to get all the stage into a place and then you can see as a modeling on it we’re modeling is a minor part of that in order to do the modeling you got to get all the different data sources put into one place in a format that you can read and reveal now the outlook for that little critter at work 14 paper sources together and bringing all the historical data is that they’ve actually got 20 setting creasing plan work completed and it’s a phenomenal amount of work but what it actually means is there’s attempts in improving productivity because they’ve got the right crews in the right place is doing the right jobs but well actually drove this was the fact that they are black book at work the maintenance which meant they never got to divide sites in time to make the SLA s so they’re okay good kicking by the government because they never got to site to fix Elaine take problems in the inside and that bad must be reduced by ninety five percent and believe me the modeling part of this is a tiny part if we hadn’t put all of Putin if all the system can’t even put place to put the data in one location work hometalk that won’t happen but yeah go back to that’s like their 85 years ago we knew there was a relationship between the number of late to the amount of work in the weather that just a cobalt to get out of doing it and this is their operational talk so once they dreamt about analyzing the historical weather danger were looking at what happens now the run in every 30 minutes and this updates every 30 minutes and tells them what they’re going to be doing so they’re modeling of the the weather the general forecast for predicting weather learning occurs of what they knew about it so all this is great you know it’s a bit of modeling what it actually means is put all the data in the right place it make some sense of it still big dates partly because we’re talking about basic and this is a another one this is a two thousand seven and big storm flooded lots of past the country but that’s the rain made up for the northern England now there are still only a couple of water companies that use rain made our forecasting in wet weather allows

depression because if you look at what happens when it rains which of these blue bars here these are the levels it all sewers what it rains the level goes up and then this bill and then we get an alarm so when this thing came through every alarm in the country went up just about and all the control rooms were in a panic because you couldn’t see the real Lance of all the fact that everything else was going off and gray and yeah it’s not really rocket science subject is available rain made out maybe 15 min it even gives you the intensity along here so you know reaches a certain intensity it’s going to go this isn’t clever no you keep getting told off by people I go if you’re saying it’s not currently isn’t and you know when it rains the levels going to go up if you’ve learned goes up here it’s going to grow up downstream and even bother it’s quite simple just for the simple things together now does that make a big difference yes it does so combine two datasources level and rain later and then looking forward in prediction you can reduce on just cancellable so old I get around it really is the fact that there’s lots of data out there from an operational point of view it tends to live in different places and because we’ve been told by actually have this vector for so long it’s just being left in different silos and and ignored so I’ll just go on to this reference hits it anywhere I’m really ought to apologize then we mailed out after talking to one of the government statisticians these said to me Cynthia anytime you see a health scare if you eat bacon you’re going to die at least 14 days early difficulty fake and etc etc etc he said if you can guarantee the sources the failure mode so I’ve got some guarantee that it’s not factual so they’ll be some hint of truth in there there might be some slight relationship between it but really you can you can’t be like I just love this like because i think you know when you talk about oil and gas it’s a very simple process you drill a hole semantics in come down fantastic and the sake of Manchester slide because it’s all been sorting see the reason for putting this up is that i was talking into an oil and gas forum and I was actually talking about Walter and the gang in the side of the road behind slow people waterfall and also my other guests and sort of the challenges they have with chemistry’s and problems and scaling it all the rest of it one of the guys in the audience look at its oil it’s really simple well water company it’s really complicated yet rains if they’ll trip pump it out you clean it but the whole point is that everybody’s industry is different we all have different challenges and we all need to deal with it but you know when I’ll say it’s got a real challenge and such a disease of always in the news you look at what’s happened the North Sea over the previous three years that’s the sort of decline in production year-on-year but there is some good news in there because that’s one of the reasons why it flattened out last year this is data that’s being generated from the UK more than just UK Phil and this is their amalgamation of data from all of the operator the nasty looking at the fall-off in production and also interestingly the drop in efficiency of the systems there’s a lot of reasons why that drop officers occurred when you do start is catered to at home when it comes out attacks this anyone’s going to secondary magnetars you need an use polymer FLOTUS all sorts of things you need to do the chemistry’s with a change they’re going to shower so you know the equipment gets all as well by the way so you haven’t got pump failures going to be issues with that so there’s a reason why it drops off a CP goes through time but the point is were here and if you look at sixty percent efficiency with the reserves we’ve got in there if you can get that back up to eighty percent which is what the challenge is the difference between that is 630 million bowel perf oil equivalents which is quite an impressive amount of oil and thought we need to do is look at how to improve that there’s lots of things that find there there’s not one thing that’s going to make a big difference to to how you improve the efficiency of em for the oil system and there’s lots of little things that you can get in there and a bit more guarded research was there a thousand and talk to 150 seen you guys in oil companies I love them in the North America Europe across the world the interesting thing is they’re all talking about the digital oil field now what is it little oil field it’s about bringing day trip to advise it and then be able to view it but it’s not that people might think about visualizing it

real-time displays understand it’s about know given the data making sense of it and using it and this to me and I think he’s quite an interesting series of the bar charts because whereas what were the benefits of getting the digital oil filter pronoun ligand and when they see the preservatives in the future increase production optimization efficiency cost reduction saves head the biggies down here real-time monitoring real-time metrics less important because if you’re analyzing the data that’s what you get in get out of it so in actual fact this here is important you do a good job so you know this will follow if you’re doing it right but to me up and that’s really quite interesting and the other thing that I think is really interesting from that server and the drunkenness the stations are followed by Garner what’s really interesting is how the change in the purchase and the application software is gone so you’ll get a slice of our politician casting at the back but this is who buys this royal field technology so ninety-one percent of it up has been bought by functional managers cyber Aaron Rodgers corporate functional managers regional manages this as much as sub but talking about Ops guys were buying this so let’s spending the money so you know for once this isn’t how I t put so this isn’t you know sort of information technology this is operations technology so they spending money on getting the information they need do a better job but equally interesting is it was maintaining it so if it’s not till become part of mainstream business and become part of the IT function it’s not going to work it’s just going to fall by the way so and I think this is really an interesting slider takes you back to the I TLT from burgers if you don’t get its converted between delivering sustainable systems and delivering the right systems then it’s not going to work well that’s just my dress with you anyway but I think one example of the win over sleep is anadarko they’re interesting totally the bromide position they’ve got 27 or 28 size across North America they’ve got over 48,000 natural gas wells the chrisley boma requisitioned all different suddenly they wanted to get an idea of how to manage this best friend and what they believed was that if you lots of 48,000 wells there’s going to be some similarities through some differences it’s within best practice so what they also need going to spend an awful lot of money on chemistry so when you do produce oil you have to treat it so if you don’t treat it it’s often more water than oil it’s got can go outside and it might have h2s in its it’s corrosive it’s got lots of scaling components in there so if you don’t use chemicals it’s going to corona scale bacteria you might even get waxes that’s probably the junior so you’ve got to use the chemistry it’s not the cheap business so frakking separate folders the tracking has been going on for generations and people just never talked about it there isn’t a lot of wealth in the world that’s not being cracked at suspect because when you produce it typically you want to fracture the formation so you’ll be all over crush it and make lots of chemicals down there so open up the formation so it’s not just something that you know what’s up now now russian oil in the world’s being practice just mostly was dating so not only concern about it yes it is so acid stimulation is another word for cracking so when well stops of presume stops producing your i sit down there when the high-pressure it’s going to crash the wells and it drove through the the formation of any sort so yes when do you do that expensive business so you know what point you go down will do scale treatments and squeezes what time you get to crackin all business to get more oil onto the ground and it’s an expensive business there’s a lot of chemistry use down there a lot of money spend with with vendors and an adapter will tell us how much but it’s a lot you know 30 to 80 million probably somewhere in that range lot more lowly ones who you the point was that they wanted to get a handle on this and also they had four major supplies that I think they thought they might not be getting a true story from the vendors of chemicals as to the effectiveness of those chemicals and when they use the right amount he’s supposed to toast at certain levels you know that tendencies also put a bit more in whether you need is over so they wanted to centralize all this veggie put any more place still an evil talking about full of terabytes over other than five six years so I’m not massive but Alex for all this and the idea was to manage chemicals to other standard costs and benchmark across the world finals good and then what is it good and

applying best practice across the whole of the fields because they believed that there would be some to margin this the margin is being far more than just the chemical saving now they learn it to between ten and fifteen percent saving on chemicals okay because you say incursions because you know what you’re using a massive single stock as well you don’t need to stop because they said last longer and you know any be needed so you’ve got a tether stuck inside but the big thing that what was the winner for them was this availability understanding the operational understanding that well in that field they’re right for the sabres are well in another field you know you’ve got the same chemistry when you’re using different chemicals people the same came to be really using different dosages so optimizing the chemicals across all their fields optimizing the way they treated them David massive increase and availability now about publishing how much is it and I know what it is because we have a little data but its its significance and this is about keeping those old world inhale so you’ve got 48,000 wells and you can keep all the booking the nut obviously because they always have issues then you know you have anymore or a lot of it and you go back to that size of the North say so it might be about looking at Cost Reduction but isn’t really you know it’s nice to have that save me it’s good to bullying the map pocket but it’s about keeping pumping and increasing availability making sure those well to available to put for longer and it’s an IT project because really if you’re going to do that you need to get all this information from SJP all this information from the production systems all this information on welfare and wealth event bring it into an area where you have to do something with it build example that’s replicated because you can’t manually configure 47,000 or 48,000 wells just doesn’t work and then work out what you want to do in it so it’s an IT project to bring the data together to then step up the operational calculations and it’s just lots of about 25 million calculations goes on every night there now this is a point I’m going to answer so what that gets out of it is real time analysis of all the costs where they spend what the breakdown is / feel what it is across the whole area etc etc etc lots of lovely analysis and how often do they do once a night because that’s good enough you know this is the whole point about real time police real time real time flavors they get information when they need it so if you manage it in the real-time environment here then you’re going to get your daily performance on target so actually we can’t respond to it faster than that so improvement to get the data first thing in the morning of mourning for yesterday is you can do something about it what it used to be was at the end of each month they get a report which might have taken two weeks to generate so six weeks out of date you can’t do anything that’s six weeks out today so one of the things that I’ll just wanted to point out is this right time doesn’t need to be second by second mini five minutes it’s whatever frequency is relevant to make it differently to the operation and quite often we see this is being hourly daily not necessarily five minutes ten minutes just it’s the right time so again daily update is enough this is the big thing improved availability wells that utilization workforce is actually for all the fracking activities so they have crews out in the field during hot tub billing and chemical treatments and really ever use that the workforce because they know when to do it but when you optimize it ok I’m going to cost one of my favorite subjects which is refrigeration and fascinating thirty percent energy use in some areas of the world so massive consumers of energy a degree centigrade cold costs three percent more so if you’re running it at five degrees and you want to drive it before with only three percent more energy bill that it would be so big incentive to get the energy management price and this is a system whereby they bringing the data after two or five minutes and then you manage that data to make sure it is tell out of kilter doing for compliance there looking all the alarms the managing these alarms the scene you know what type of calamities and the kingdom on ears is managing alarms as in making sure it stays an alarm for long enough to respond to it because if you move the chicken around in the fridge and block the vent alarm go off I’m guessing about private outfit which types of all the chicken is kind of forget so understanding what those allowed to be making sure condition for longer s’okay it’s not clever it’s just about understanding how the system operates and how it works and what they get out of this is the ability there was they’ve understood it to go and control it so this isn’t looking at all your set place so now they’ve managed to set points the optimizer site they record what the optimal performances and then they monetary if Sonny goes and changes it they change your patent rights report because that’s your optimum efficient operating point

so this is all controlled remotely the next stage for this is we don’t do this there’s lots of all this they do little chilly so now because they’ve got all this real-time access to data and these guys are real time they can actually have a contract with the electricity supply it to load shape so st. waves in the u.s. they were going to call to say you need to share too big for four and a half hours yeah two o’clock in the afternoons at six thirty because we’ve got to come with the the amount of power cord analyzed let them do that the long check to get a load on these sites running the mish mish matrix raise the internal temperature of store turn off campus all these simple things first then start raising temperatures in cabinet help deliver to keep saving everything’s happening and of course what they get out of that is a big saving on their sanity so what do they get from running refrigeration management systems these are big numbers twelve senergy said because they’re running it more efficiently there the right savages obviously regulatory compliance is the whole thing about monitoring them is you’ve got to keep your fridges at a certain temperature and your freezers at a certain temperature if you don’t you in trouble and you need to be able to keep State seven years to prove that you’ve been being good at it but that’s the main drag across regulated compliance always is the supplementary ones are seven percent reduction alarms thirty percent reduction in call-outs so 30 less fans on the road driving brown if you go on the motorway next couple days now I’ve told you this you’ll see white bands with snow claims on the side there’s over two thousand driving around the UK fixing fridges and freezers in supermarkets at any point in time is about three Tesco hypermarkets if you worked out the number of cabinets across the whole sites not selling food because the cabinets are down so it’s big so yeah big enough sleep name is plus lots more so keeping some customers happening Nick Tesco’s probably neither at the moment and one of those strange things if you want something efficiently you put their stress on it if you put that stress latexes really efficiently it’s going to save more pebbles and use less energy but more importantly what breakfast often so the mean time between failures gets extended and more important than that because you know monitoring the types of asset you know how long it’ll last you know what the problems are going to be with that type of asset so you can find the access that I longer different manufactures so it’s just about bringing that data replacing and analyzing it really and this is our perfectly fly through this one this is gone and determined this is probably one of the best examples of operational analytics than I like it because of where it started and where it got to so South Australia water areum Adelaide as with all these things and they always run out of water so we need do something about it the business of impotence our nation plug so they could produce fresh water preferably then realize because the thought you to run it so they need to manage how he did this the work they did this was to use the existing systems so what they had to do was demand forecast if you gonna produce the water that people need you need to know what they need so first thing to do was to work out what the demand was going to be so they have a weather forecast they have all the historical data they know what’s being used in the past they know what plants are making and surly these Englishmen forecasting and the mother model which tells them you know what was they didn’t even wearing and needing to do that they just have to link into existing systems it’s already there it’s already been put in place you’ve got me telemetry system which is no scientists and ability customer building because obviously appspot gives you money so you can have a building system you’re going to mex obviously because everywhere for the next office they have to do all the water quality because the regulated organization they’ve got Maintenance Act data and activity in maximum so they’ve got a 3gs spatial system wasn’t anything that was missing its all there they all exist if it just wasn’t being used in a way to manage the system’s life so all the day built a model predict the demand across the whole network this is a nice one well my simple model is just regression analysis looking at historical data and and what you do is predict what the weather’s going to do what the demand to intervene and from that they can work we’re ending the water so well is there is the relative man forecast they then pass that out to a hydrogenic modeling tool which will work out the production requirements cloud will tell them these scenarios were also these producing the various facilities to become demand and that will also do on the basis of an optimizer so this is you see here this is all the different sectors like customer sector by education hospitals Joan roads all the industrial different categories of the different lawsuits non-revenue water velocities birthday so

if he works out whether it use the water then pass that out to an optivisor which they already have and the optimizer then looks at the best way of making that water with the lowest transportation calcinha lowers production costs so now they’ve gone from demand to production planning to optimizing because they now optimized that that production of this here you can see is the brown linings of the next 24 hours and this is there seven days because they know what they’re going to need the next 24 hours in seven days they can produce it at the lowest cost which means that the lowest tariff so now they can actually put in it the deal with the logo electricity supplies and then grabs updated by the opportunity on spot market so they put in a complete link for cats will be going to need optimize the way you going to produce and where you going to produce it and what you can to run and then go and buy the electricity of the cheapest price all of a little bit of modeling along the way you gain all sorts of advantages like because you’ve got linked into the control system you can see what the churning the tank is and where the files are opening so it’s all engineering stuff we can see where the watery the reservoirs are doing because the data is already there so these are just a dolphin’s these are just nice to have you know you can link it into to the acetate and click on it and it’ll tell you when you’re producing but the real thing is that you sadly the weather forecast you endless doing demand forecasting from that you production plan your then look at where you need to store it and transportation costs you then make you for the cheeks time of the day so there’s lots of course produced overnight and then pumped up to storage you then know your energies in the bc you have advising spot market and you can also change the way you operate the site so I belated strange place December 2013 for Cassidy was going to reach 47 maybe even higher degrees great so they knew that the spot market which is going to go crazy and they’re all a variable towers down there is energy you simply man so what we live is he went into management production they filled everything they have and they’re going to draw down for the following week and they just drew down the water is positive media they saved over 1.3 million dollars in a week just because they could forecast what was going to be needed now point is they haven’t changed anything that they have it’s the same equipment as to say production facilities it’s the same control system everything to say it’s just that they’re doing a little bit of forecast English bleeds on and cascades down to overall efficiencies as with all these clients they were all telling what they’ve actually saved look at how it is circuit 10 million plus dollars a year just by looking at weather forecasting so what kind of Martians really around that forecast I can sing some sighs we have all sorts of different variables they report we have handling on there no they’ve been released in two years now and the last point there is they going to automated control during the course of this year probably the next year so for the moments they wouldn’t want let me half hour and I’ve using the production plant which they might near the internet so they provides the control systems and they’ve updated every half hour that there’s all sorts of checked in there and balances and you know we do a lot of blood of checking on it in but they’ve got enough confidence now to go to automatic control so they’re putting the system in knowledge from their production plan from them all until he’s going to lead to automating controllers production facilities going into next year so they’ve got confidence is so I’ve asked the only thing in Satan’s if they got confidence from from running it I didn’t want to bore you with with PowerPoint but there’s lots of slides will show how they can use the model there’s various different models they can apply they should go back to the slide back here they have up here they have scenario planners so they can actually plug in different scenarios and say look you know we think that I was an outage on these sites we need to do some work put that in reading on the model so anytime you can run these scenarios and the old slaughters is options within the system and you know they can run the numerous models for different time periods and looking as they go in the futures 22 how successful it was but the other day the proof is in the eating and they’re going to go to ultimately control so i think what loads of avoid south australia for this they just want a globe award in the water industry for running this system so what are they got absolute well money efficiency availability they know about the estates they have a real good view of where they if they are the big thing is the library of desalination is that within the water much better so notice about operations they’re very happy oh yeah is happening there’s no doubt at all that operational analytics industrialist whatever you want to call it it’s happening it’s been

about subway you know years ago you would see NASA mind you know let alone operational analytics now you go and see any ice advantage you they’re talking about it they want to talk about it we got the confidence three years ago thousand IT people two years ago maybe 150 people from operational background last year one in London 1500 people at least the operations guys so it is happening I say so cool because a lot of the stuff you can do the easy wins out straight forward it’s going to become more advanced so we’re going to get real-time modeling we can get a lot more hour I things like this really in real time so we can actually do some advanced modeling it and understand what’s going on I still leaves a lot to be gained by doing the simple stuff first and I think there will be common purpose grows out here and the last point of ice and wages i worked with yorkshire water when they run out of water 1996a very friendly new members they had 750 tankers going up and down again while they’re taking water from Keeler’s appointments decisions there’s no inter connective is he the Chairman got sacked we’re going to his mother’s in Lancashire for a shower at the time you 1996 they were ranked 10 out of 10 for the water companies they were the worst that every measure the government used to measure the effectiveness of article 10 out of 10 they put in a system called shortfalls asset shortfall they wanted to understand how to be proven how to do you know drive the business better they put it a comp like I tott so they got the operations guys at the IT guys you can’t say no and you can’t swear so sort it out get together and build systems at work and in five years they went from being the worst water company to the top water coming in all the world category along the way nice a 47 in group or eighteen percent of the objects every year which is quite oppressive and how do they do identify using the IT guys working alongside the LT place and you know they were probably one of the first to really understand that and start to work you know the eye sees only good if it’s delivering more the OT guys one and from the same thing the obviously you guys got to actually stay what they need otherwise not gonna get it when I come back to the guy from konechno you know protecting earth you’ve been tough to say no before people internet now is going to say yes and find people to say yes to so it is changing it’s definitely changing thank you for the quick answer to accept a deal vanilla come on up son ascend silence i’ll start off with when the air is under increasing interested in is about the data security particular round security of an instructor what are the risks of only on the desert the operational analytics running the risk of also potentially exposing later that you might not want to expose how’s that thing balance um it’s always going to be an issue i would have saved much refrigeration i think that probably one of the most security conscious companies i’ve ever come across to walmart so it’s not your classic you know a lot of companies or the utility companies its retail that it really really good and yet they both at so then their setup would actually now they’re happy to go that way we’ve got systems which measures the five minute periods all of the oil flow down link the network east coast of America all of the oil companies got access to that and there’s a there’s a contract script with the company that just as penetration test continuously what can we do now they love the data they benefit from the data and we try and make your secure as possible but it is a trader but I mean if you look at the big oil companies it was always inside the firewall you could only ever thought server installations in there changing you know I need some changing rapidly you know it’s so years ago you couldn’t do anything outside the internet wasn’t in half and you were going to get web access but if you didn’t you don’t know how can a lot of companies which in the last two years have changed our approach to to that now it’s the same approach engineer so our cybersecurity is without therefore it’s happening you know it’s and it’s going to have to happen the other thing is giving people information and he don’t give them access the data so you know there’s a novel aqua driven my life but you don’t need to get Matusz the daters giving the pictures tell about to do given the descriptive you know he’s don’t do this and it’s what you need so but i won’t give them that no question

but i think these were all grades are really shots only I working we have many ideas of what we what’s your advice on and how to prioritize how do you spot the golden eyes I think it’s always a case if you need to engage with the operators you know because they know how they’ve running I where the job in come but what are coming in Northwest and we did this lovely bit of work and research them you’re spending a hundred eighty thousand times a year too much of total we know well the operators knew already but nobody never asked them know what what did they sing they could do to improve on it the answer was changed these contractors that keep fender so our point is always driving games with the operators engage with the people and literally start with it with a whiteboard now if it was possible what would you like to do well then let’s go and work out what takes is available but today we can use and which of those we can deliver to you never run before you can walk start small and it’s always see success driven success it’s a lot easier to do that which also helps you because you only need a couple of data source if you have and everything to what extend to these opportunities that you’ve proven or confirmed using using data versus opportunities that you’ve discovered by our metrics on the data we very much come from the you know what what’s happening nowhere to kick it for you know the former so now we quite we don’t do any advanced analytics we do some are modeling with you sir but we don’t do discovery what we tend to do is work with companies and said well you going to the discovery and then won’t we find willing bed into the operations because you ate it’s about understanding that there’s no such thing as original set it’s happened before it happens tab and sour kick so capture why it happens and embed that in the suffering of people in early morning no so integrity operating windows telling them when they’re bringing outside of agent and giving them the guidance as to what’s going to go wrong if you keep doing that is the note from our point of view it’s not about discovery now you find lots of things along the way and I don’t mean to belittle is but we worked in curry factory one of the things we found was that the reason why the chillers kept icing up and they kept breaking was because somebody kept probably the back door open the GST Nick carries out the back door so yeah so you do find strange things but no mark point of view what we do it’s much more about embedding rules and understanding okay so actually you know the opportunity is there that you’re using the IT normal dormice discovery in time to actually do something about it can capture opportunity yes yes it’s there it’s happening you know I come back to the oil system we’ve now got so much data from 28 fields and 48,000 wells over there next week that they want to do a text discovery project because they’ve learnt a lot they save a lot they’ve improved things a lot but they know that there’s got to be a lot of information in that data as you know a lot of information which we don’t know about so they want to do a discovery project to see if they can you know finding relationships which they’ll then embed us rules in the software got one more question five committed a different time the way you talking earlier about your collecting nature in real time but you’re analyzing and looking at it in longer periods an hour a day whatever as such sounds like you probably is using through off-the-shelf technology for for storing movement so young nothing special nothing special your analytics are properly specialized but your google actually means of collecting the day turns the page repositories are probably just off the shelf signal yeah you know I’m going over the years and views Oracle and various other things for the mature all of us people now you know we are talking to some of the big debut think they’re people because they’ve got for being a lot of these data collection systems are out there now there’s been big data so okay but there’s going to be born dead sorry so we might be changing for know from our point of view we’re collecting data five minute you know analyzing it five minute and some of it quite deliberately we might be picking up about say to every two minute but we only through salereal eat every hour because that’s the response time you know it’s you know that’s timely enough yeah so now we’ve been dating we analyze it as it comes in or we might match data and the other thing that this is really quite key to where we chose a lot of the data we’re going to pick up might be coming from allow it might be coming from outside reading so you’re not going to get it online you know sort of the coming six on the web fall and then when it hits the database that’s when you want to do all the calculations so as soon as you get it is good enough we know in the real world of this data is collected manually is collected on tablets or used to being between notebooks now you might be collecting data in control systems every five minutes when you wait for

that tip 10 really good thing really like a retro for you to do something with it there’s still a lot of good stuff you can do with it just my opinion but when i kin okie integration motion from comes with a bigger association because on TMO sleep out dentists and i think i’m quite bright black Irish so when you’re when you’re trying to quaker the first issue in tacos a equation association datasets so I mean you didn’t mention anything did you have any knowledge to share yeah I mean the thing is that well the big assignments is you know what are you going to use is a master data registry now from our point of view we normally go with what the acid registers so you know when don’t like putting a system where is he that make me smile insist or there’s a control so Paul is that an essay to the asset register so that would be asked that point for the point is that that might not be the right structure for you to do the analysis so that’s why I mean about structure doesn’t you might want to take that fun and read it has regroup I’ve come to know groups but you’ve got to have one master record you’ve got something at all the limbs tibias and when you do need approach so when you do the integration do you do ESL one opportunity actually you be doing inside a upgrade sort of program where once you’ve done is Association danger and then definitely new consistent important place where I people working together or we’re just doing it for time is good because money because otherwise it just becomes so outdated quickly so you know we’ve got these I’m application and updating so that you know you keep the things this is think you can look after a few well you can look after maybe proper processes you know if using an Excel spreadsheet you can’t do that when it gets beyond that and the other thing is it it’s going to be everything in one central repository because you know otherwise create Express language in HEK cells personal opinion you know once he gets into Excel is just personal opinions got to be more source of place you might write an X up report that’s fine but you know you’ve got it ultimate those legs because otherwise there’s no trust in a valiant effective so yes that replication process is this kid putting the legs put in a mapping and then make sure it is it automates that’s Infante only as you probably know the soundings crease production in order to collect bring the price down to try and take on the American Franklin which works as which to to extract oil in the Middle East but it’s backfired because about shot a few sonic and a red but the two main silent and increased production the fracas America Halliburton massive increases in productivity which is which is a problem is the seventies and everybody else I’m sure a lot office cuz you know due to the techniques you’re talking about fear the the fracking and the shale gas in moments comes in different business it’s not like the traditional oil gas versus I mean a lot of the nations are shooting in a different business unit is different you know this is you almost got as much in water treatments of back end of its you hang on on oil technology the shallow they’re different crappy but you know it depends if you talk to people your gasoline whole thing about sound is dropping the price the Americans suffering ground you know there’s a footings for planning more ice and complaining a lot more so it’s sad i don’t know why there’s that isn’t done it or how long he’s gonna curve ball but yes it’s an interesting thing we’ll just another contrast and also because people want to improve efficiency other processes because weather sun hundred dollars of how segment of the dolls about is really nice but we’re using an ama commentators are saying has often been any increase in productivity the recession period and the store going up now and they say they’re all waiting for robotics to come out that’s the next new technology new wave of technology but for why you’re saying I think there’s a lot of productivity to be bought on existing systems if they can’t actually I think God is is getting into the service industry because mainly this is chemical manufacturing industries or you can see what sacrifice I think it’s much harder to get efficiencies improvements in service tonic industries which most Western governments is driven bali you’ve got to be a desired that’s the thing that the key thing we work with company that’s the drugs that incriminated efficiency in their filters we might do neurology installation and that’s no there’s lots of people doing even a design design anything you can do something but I comebacks appointment sir you know when you look at this fifty dollars power imagining there is tiny when you look it up under dollars

apparently margins pretty significant because if you still actually still operated 50 than you mean we were making quite a bit of money so I think the whole drive through efficiency improvements and you know the last dip 2007-2008 there was a directive of action brought in them because of we need to be more efficient so yes it’s going to drive efficiency improvements because then he’d stick would you say the data so it’s just a matter of logic there question about happier problems many good enough data to drive I come from operational backgrounds the data’s of it I’m a model and I’m build models like that and not being able to get places to do some good yeah when data is not use the data quality is always forward when you start using data type of quality improves even because you know it’s wrong or because people take more care with would normally do and calibration you know there isn’t a single project we’ve worked on where data quality doesn’t improve as people realized using the results so offshore technicians will fill out anything and then when they realize people actually take note of it and using it kind of walking further I think that no it’s always the text but there’s lots of filtering you know I mean we work on the National Grid project and first thing you have to do is you’ve got to filter the data you know because there are always altitudes and you’ll get an instrumentation 0 you know so you’ll get a full scale of zero step you’ve got a lot of those now if you want to use the data but we have all sorts of tools for averaging filtering counting values you know because even if you only get 23 readings in a day it’s an hourly value you need to know that you only get a few people do you drop a future in the course of an hour you need to know about how do you feel how do you deal with how did you ignore them so there’s lots of approaches but yeah absolutely right you know the better the quality that the more sense you can make on it but I come back to it and that bottom on there y’all should watch is a prime example the texture quality of your water improved matically once they started using the data because if an asset manager is responsible for a water treatment works and he looks rubbish then we’ll find out why Lords rubbish and if it’s down to calibration of instrumentation down so the equipment failures put it right so okay any more questions the deal well thank you I guess an observation and so I what you need oil gas industry myself and we’ve had many real time or near real time value an existence or the set of a project controllers from synchronize it will be embarking you know what these systems but very much regionally deployed in some we operate about 15 regions monitoring real-time production for officiating all the big challenges of any interesting thing which rotation as pleasure sweetie I to notice that most of those systems of which render service for life now and the way the technologies being driven we’re looking to centralize a long will you keep you which means streaming data lacking in near real-time solicitor fees minutes hours day some of its kind of 10 second leg streaming data up in real time and then I think that the Prudential strong each other and networks you need to score that the quite a doing mode well the last fight about is the Australian firm that as you start to move towards close with control so you’re actually starting to make mission critical decisions with data on the corporate networks it does again really challenge because we got so he’s a week for very much precious control mission critical systems say our own work patrolman lies behind EMTs and there are providers around responsible actors trying to compromise Prost control systems as a rigorous and really digital security Lee the question about trying to align the logical integrity of data as well across the culprit operations massive challenge but it’s huge announcement happen yeah also interesting thing is the way the oil and gas companies are changing and we were the company that’s a wanting to implement systems across all their assets but they were locally made so an asset was was king and they’re changing that so for a while we were the recommended solution and next year it’s mandatory so they decentralized so whenever the gas has traditionally been fairly dispersed with local control tasks and do what they want within reason we’re seeing that tonight a lot of what the wrong thing is yeah I think there’s a lot more focus on studying evaluations yeah the ethnic diversity

right up in week the other thing is you’ve got a lot of history on there you’ve got some all that sex even in the North Sea they’ve got a really poor comms and if you look at South Adelaide there they were sitting so you know they’ve got five but they’ve got good columns so it’s not the same it’s not as difficult for them you know they’re not been both sites so it’s different challenges I think is very similar tragedy how much yes we want centralized computer and also if it needs a push some of the and let it out to the edge and that’s what I’m doing gia dinh what we listen to that yeah how you actually federated analytics oh yes you’ve got the central computer but also pushing stuff down to machine also the machines to make choices as well yeah thoughts of 70 Federation you know you’ve got any controls it really powerful so what you learn can be put down the head end so that those loans you saw there were also suppressed by the controllers actually install you know and they’ll send a message out to you know to local people so you have to get us the store manager get a text assemble through so different business but same challenges well we see a lot of that you know local processing local servers you know just dealing with different regions and then all the data comes up for central reporting so you’re sending a lot less data central authority but the exit is none of the sites ok I think of a calm down more questions for now come on take a break here not to carry on like a cup size of a nail and discuss a colleague setting any views on that they’ve got up like to join me thanking you and the different VCS