Simulation: Catapult Exercise - Round 1
So our exercise and I was going to be walking through it and then we're going to organize a little team up here and like I said everybody's going to get three shots so our catapult here is uh already preset for you the pen heights and so we don't have to worry about the pen sites pen heights up there set but we do have to pull this down so that's tight from this little knob so we've got some tension here all right? And it says pull the arm backto one seventy seven you'll see there's a little protractor here on the catapult one seventy seven is a little pointer one seventy seven is almost all the way back but but not quite so it's almost all the way back so we're gonna pull it back to one seventy seven and shoot the ball now make sure you keep safety in mind with this because sometimes you know the catapults positioned up there and people just get a little too close uh don't shoot ball on your own here all right? Some people might laugh but yeah, it can hurt so we're going to have someb...
ody measured the distance of each shot so the catapults going to be on this little stand we're all gonna take a shot we're gonna have the uh a measuring device take measuring device and we're just gonna put this right out on the floor we'll shoot the ball over this way it should it should go to about here so we're just going to set this down now the other thing is um try not toe step on the tape measure or move it because that's going obviously introduce variation into our process so while one shooter's shooting what I'd like to have his two team members standing you know roughly right up here and uh identifying where the ball lands on the first bounce so it's where the ball lands on the first bounce that's the that's the data we want so we're going to capture that and I'll write it down so I'll capture the data up here for the distance okay and each person's going to shoot the ball three times and between shots we've got this policy and uh this procedure so between shots if I shoot the ball all right between shots I loaded I shoot it I have to remove the rubber band and re tighten it while I'm doing that somebody's retrieving the ball for me two people are identifying where landed one person means right now recording it so we're all going to do this and then when I'm done shooting alright my three shots I will change with somebody and they will come over and shoot three shots so that everybody gets to shoot three shots were going to collect the data and see what the variation looks like all right uh now we have attacked time here too in our tak time is fifteen seconds what does that mean? That means we need to be getting a shot off every fifteen seconds just like in the lean sigmund game we were getting a a sheet off every twenty seconds. So fifteen second tak time that's our drumbeat will record the distance is and then we're going to calculate the range range is the distance between the longest and the shortest? Now what happens if somebody gets up here and they load the ball and then they pull it back and pull ups that's a zero no mulligans do you know what that does to arrange that would make our range enormous because now we have a zero to whatever the longest shot this so we have to be very careful not to drop the ball so it's like that line and uh the movie stripes with bill murray's asking where's the sergeant he's blown up sir, don't blow yourself up with the with the ball here by the way we run this exhibit we've run this exercise in the military for a number of years in the defense industry and whatnot they love it because it represents, you know, shooting stuff but where I want to blow ourselves up with the ball all right so that's going to be that's basically the exercise and then like I said, I will I'll capture the data while we do that so I just need everybody to think about now ok let's go back over the procedure shoot three shots pull it backto one seventy seven let it go don't drop the ball, remove the rubber band put it back on and we get situated so we're just going to play with play with this exercise to capture current state data. Any questions on that we all good okay, so, uh why don't we uh curt you want to shoot first and while you're shooting, why don't we have kate you and jane be uh measuring where's the ball gonna land I'll give the ball to kurt all right? And, uh I will record susan, how about if you retrieve the ball and round while kurt shooting incidentally, when you're done uh uh let's have ah let's have cat follow kurt all right. And kimberly follow cat in the shooting kurt when you're done, you go retrieve the ball susan, you can you can get in line, all right? And we'll just kind of we'll rotate this way and then we'll we'll get a couple more people measuring so the case so that you and jane khun get get shots so we're going to get, uh, get six people three to three shots each we're going to get eighteen data points all right and uh and have a little hopefully have a little fun here this fun tow fun to play with all right kurt you could start shoot whenever you want remember we're on a fifteen second tak time so we wantto keep it going backto one seventy seven did you guys see where that landed? One fifty one seventeen I'll give you one sixteen go again no you know it one eighteen all they want us to move around okay I don't get to reconnect the rubber band took it off reconnected I'm left handed one thirteen dwight take the rubber banana within and between you in between shots go yeah I want to know how you transferred all of these with southwest airlines you can bring them with you on the plane e didn't distance every time ninety three ninety eight which one was anybody hit the same no it's being different every time no one said the same one yet yeah so far we've got a range from ninety five toe one twenty I like how everyone take me I'm gonna call this the one me method here kimberly crouch system seems to have a little variability and that the best one yet well camera lee there take that off variation has a funny way of creeping in the most interesting places but thank you inspired susan here really some similarities I would so you know it's pissed someone on one o nine okay you know a was it right there it's just a fun part so I got you I got to win votes e I think that was one e one twenty to teo green one twenty four did you do differently that time all right so yahoos who hasn't shot you have kate your skates shooting without measure because you don't like ok you have todo cat sneaking around you're going out probably actually I hit him probably hiding around dad's gonna have to work with these things really what does he do? Can I get a rubber band before the ball thank you e one twenty two wow did you do anything differently that day on that last one I didn't think so didn't seem to be all right is everybody shot she did james gotta shoot you chain all right okay you want to help just uh help kurt measure we'll get this too couple of points here ah come on hee hee hee all right very good. So we've collected our data I think everybody has shot three shots correct okay, thank you very much so what we just did is we uh we've armed ourselves with some facts we've gone out we've collected some data way said alright let's listen to the voice of our process and let's observe what it's telling us so um we might say oh yeah yeah we can shoot we can meet that deadline we could hit that price we can deliver on that date, whatever the case might be, but do we know with some degree of certainty and confidence, or we're guessing so this is about knowing, not guessing this is about shifting from I think, too. I know, by the way, I had a ceo of a hospital tell me years ago, john, we're just trying to get from I hope too. I think, uh, I can't make this stuff up. I was like, really, like, yeah. All right, then we better think prevention when it comes to health, right? Because, uh, hopes not a good strategy. All right, so we've we've collected our data now, and, um well, let me, uh, let me just capture the range real quick. So one of the most common measures in statistics is range range is a measure of variations, simple difference between longest and shorts, and nobody dropped the ball. So that was a good thing. We were very careful about that. Looks like our shortest. We've got a a ninety five years at the shortest. I think it looks like it, and our longest was the one twenty four. So we've got a twenty nine inch range, all right, twenty, so we have a twenty nine inch range. We could calculate other statistics from this is well, we could say well, what's our average sometimes we've heard of that is the x bar the y bar what's the average you know add up all these numbers and divide by the this case we had eighteen shots to get an average shot, so on average you know we shoot the ball, you know, whatever that might turn out to be a hundred ten inches but think about statistics his average is never happened because there'd be decimal points too probably we're always landed on one side of the average of the other so one of the lessons we learned with six segments that customers feel variation they don't feel averages always say this is like the pilot coming on after you've mr connecting flight and you're three hours late and they go on average we're only five minutes late you don't care you just missed your flight so averages you know we use them a lot to measure performance in business but we also have to measure range or variation now another measure variation would be things like standard deviation all right, so you're measuring range in different ways of measuring variation in different ways but ranges pretty simple I think it's probably about fifth grade so anyone you know with a little mathematical sense can do it so now we go you know what um I don't john is that good? Is it bad? How do we how do we know it is a twenty nine inch range? Is it good way pretty good in our business. So without context without comparison without benchmark data without clear understanding of voice of customer truth is we don't know what if the customer says you know I need you I need the balto land within ah with within a forty inch range well, we're not we're not bad but the customer says I need the ball to land within a six and strange oh it's like I got most of the car in the garage uh that doesn't always work. So the idea here is to compare voice of process that would be the car with voice of customer which would be the garage uh we take in the sides off the mirrors off we get it in there without any risk. All right, that's what we're gonna play with here so we have to check out I do have to define our process. Could we mapped this process? Sure, we could step one what load the ball step to pull it back step three make sure you're at one seventy seven step for letting go, you know we can we can map it out, we'll do that we can calculate that x bar that average on the range calculate our standard deviation we can calculate our voice of process can calculate our voice of customer we could talk to the customers say what do you need? All right, I need you to hit this target I need you to hit this date I needed teo to meet these specifications. Okay, we've got the voice of customer that we calculate six sigma measures like cp and cpk which is essentially process capability so we can we can actually calculated a couple different ways uh and populate them populate with that data are measures and measure our performance and things like that and then take that into our analysis to our cause and effect and this is essentially where we're going to be going um with this exercise alright to learn more about variation so mapping the current state so this would be like, um no let's uh let's pull out our define tool from domestic and map the current state and a simple illustration of that we'll just do right now definitely a lot of organizations khun spend weeks and weeks and weeks and perhaps even months mapping things uh, my experience is that it really doesn't uh take much time at all teo to map a lot of processes we get the right people together we use post its and wallpaper to do it we don't sit and spend lots of time on video initially later on if we want to fancy it up and make it look good to distribute and publish that's that's fine um but for the purposes of mapping get people on their feet I know what you're doing know how to use the tool but say all right well let's uh let's walk through this process together this is part of that alignment process right up front it's so important so all right um first thing we want to do is, uh identify what happens right up front and I'm just going to call the first step here because we sort of did it for you set up so we we set up the catapult and we set up the tape measure way set up the operation okay so set up and then set up goes to now it's time to shoot the ball all right, so let's uh let's shoot the ball or start would start preparing to shoot the ball so what do we do? We're all set up kurt you were first what what what was next already set for me so I didn't need to do that so is basically placed the ball in the cup okay, so I just got load the ball all right, so there's a step um ok good then what do we do then I pulled the lever back okay, so we we pull back okay then what check for a hundred and seventy seven degrees right yeah now a lot of times on a on a process map we get to one of these decision points you can see it represented here by a little diamond and so we'll throw a decision point on here she was a different color for it too and we said we pull it back to a decision point that decision point happens that's a question mark usually in this case is at one hundred seventy seven degrees all right and if the answer is yes uh what do we do okay let it go all right so we release it we let it go and if it's no we got to go back and adjust our pull back right till we get it till we get this to be yes and then we let it go sooner or later we'll let it go all right what do we do after we let it go just come back this way we could map it out any way we want to measure all right now we measure it okay wade oh retweet okay incidentally while we were measuring I was also documenting so I was just captured that and then we had somebody uh retrieving the ball yeah after that for reset so I'm gonna go way retrieve the ball all right and return it okay. And meanwhile while that's happening is part of the setup the operator in this case is uh taking the rubber band off and putting it back on again there's an assumption there that maybe that has something to do with the elasticity of the rubber band so we release it and we put it back um it's going to give us a more consistent distance now that's an assumption okay that's so it's been designed into the process but we don't know whether or not that's a fact or not but it's something took something to consider that's where that take it off and put it back on because it seems like that's a crazy activity it's redundant thing does that matter? Well the truth is it didn't seem to help much but that doesn't mean in isolation there are other things you know when we look at it outside of the isolated incidents there's a whole lot of other stuff going on so one of the reasons we process map things and incident how long did it take us to do that like five minutes at the most so process mapping again this is a simple little process but we can process map just about anything in a relatively short period of time if we have the experts helping us walk me through this process okay and it's a brilliant way to get people cross functionally to look at things from a more cross functional systems perspective because a lot of us just don't know what's going on outside our own little world so process mapping allows us to see something beyond her own world and the interaction of it so process map whatever business you're in all right, how do you go about selling whatever it is you sell and then delivering it take it right take it right on through uh value stream alright and, uh color coded if you want but keep it relatively simple now the idea here is that there's an opportunity for variation to creep into this system where everywhere everywhere could we have variation creeping into our set up hint what's this all about um we have we have opportunity for variation right here. What about loading the ball? Is it possible we have a variation, you know, creeping into the way we load the ball and I'll give you a hint? Maybe maybe that has something to do it I want to hurt myself uh that that could have something to do with it. So the way we load the ball while we certainly don't want to drop it, okay? So if we put that ball in there good and tight you know that's going to affect the distance we have more variation creeping in what about our pull back this whole equation going on over there, any variation going on there? Things are very shady things going on the iie mild, some of you you know I heard a couple of times while the seventeen wasn't being followed and I heard a couple times from our commentary hey, good shot well what's a good shot in this game that way everyone pizza party okay, but it's funny because what happens is we get somebody that's thinking distance maybe no maybe like the longest drive in golf if I can really get it out there boy I'm gonna uh I'm gonna hit a home run so uh you get people that will do something like this will get it back to that one seventy seven and then it's you know there goes a nice one good shot so that set and I, uh saw susan do that once just knocking it out of the park it was funny it was there's actually when you go back to data susan you were like what may be the fourth shooter yeah, you've got the long drive but you were real consistent here that it was just I'm gonna let this one go way so we think you know what? We've got variation going on here we might have variation going on with our release method you got the thumb you got the two finger you got the one finger they've got the different release methods if you noticed but really gamba and you go see where the work is done and you watch we've got variation creeping in here now, was there any variation in our measurement system? You know, uh, let's, just go with it because I was standing here watching, too. And I was listening and documenting this so that we had all kinds of variations creeping in our measurement system. Now, I think I I documented what you told me, but there could have been variation in the recording and reporting for the interpretation of the reporting. You know, not everybody interprets a report the same way, so there's variation that can creep in there. All right, we could have variation. Uh, in terms of ah, take it. The time it takes to retrieve the ball, get it back. It could have affected our tak time. I wasn't capturing the the tack time in detail, but we could add variation around that as well. So they put the point of of the mapping from this particular perspectives to say, well, we've got a lot of opportunity for error here, right? We've got a lot of opportunity for or poor quality. All right, so way. Gather all of that from our process mapping activity. And then, you know, from the data we can start to also define and measure things like our range. Like our average, you know, this is again basics statistic so hard, our longest from our shortest this illustration this has to do with the time it takes to do something from seven to three hours. Incidentally, if our ranges it takes from seven to three hours and somebody says, how long is it going to take? What do we tell him? This is like calling up a company to come and service something at your house and they go well, okay, you need to be home, and we'll be there somewhere between eight am and eight pm. You know, on a tuesday when you're supposed to be at work, how do you know anybody to do with that right now? Try that with lexus, they, uh they'll ask you what time, and I've got friends that have done this, uh, I haven't tried it on lexis yet. Put good friends that say, I want you to be there at seven fifty nine a m just play with seven fifty nine or eight twelve some random time, you know, we're used to the paradigm of eight o'clock or eight thirty they'll knock at the door eight, seven, fifty nine or twelve it's hilarious, they're really that precise about it, okay, so we could calculate our statistics here we can calculate our standard deviation incidentally uh the days of manually hand hand calculating this stuff which which was quite scary to a lot of people now it's a simple is loaded in a spreadsheet and a few commands and you get your get your data pretty quick all right so we can calculate our standard deviation which is uh essentially the square root of something we call variants all right incidentally it's also uh on a graph on a visual aid if this is a district normal distribution curve right here one standard deviation visually represents the distance from the mean toe where the curve changes directions where it goes from this direction to this direction that's called a point of inflection and that's where the that's what that the measure of standard deviations from the mean to that point so that'll come in handy a in a few minutes all right and this is process variance again not to get too scary we don't have to manually calculate these things anymore we just loaded into a software it punches it right out force but it tells us variants standard deviation and range are all measures of variability so what we really need to so whatever measure we use we need to know um how much do we vary all right now this is an interesting little, uh chart because what this tells us is that our true voice of process all right is going teo fall within this distribution all right, so we have ah, when we we lay this out we have our average plus you're minus one standard deviation plus you're minus two standard deviations plus or minus three standard deviations and statistically we can predict with confidence that sixty eight percent are going to fall within the average plus or minus one standard deviation it's just that's gonna happen ninety five's gonna fall within plus or minus two standard deviations in over ninety nine percent will fall within plus or minus three standard deviations we know that so we essentially no but with a great deal of confidence that in our process all right ninety nine percent of our shots are going to fall within this within this range roughly as long as we have enough data points but actually we have eighteen data points here we should have a least twenty preferably thirty two have a really good, reliable prediction. So we have quite that many, but the idea is if we have enough data and it's it's just it is what it is we can project dick reliably that we're going to fall within this range now that's his voice of process so john, we're pretty confident we're willing to bet good money we can fall within this twenty nine inch range, okay, but I need it in six that's a whole different ball game this is voice of process this has nothing to do yet with voice of customer all right, so now we start talking about voice a customer this's the garage we gotta fit the car in this is the goal post these air the goalposts so are slower spec limit and are up respect limit let's just call that a twelve inch range so we got a twelve inch wide garage and a twenty nine inch wide car you're uncomfortable that process probably not right? So let's leave the car in the driveway. So, uh, centering is not even an issue in this case are our process is just way outside our our boundaries. We're just not going to compete, so we've got to figure out a way to get that variation out. So what we're now doing with process capability, things like cp and cpk is we're comparing our voice of process with voice of of customer so if this is our voice of process and these are our goal posts, okay, so we've got our spec whip divided by our process with well, if our suspect with is twelve inches in our process with his twenty nine inches, we have a problem. I have a big problem, all right? And cp doesn't really consider centering it doesn't it's just it's looking at potential all right, so the potential is, um we've got a twelve inch range if we were at six signal by the way we'd have a ratio of two to one we have a process of a customer with of twelve inches and a process with of six think of a basketball in a basketball rim it's a two to one ratio that's actually six sigma potential ratio so we've got a a rim and we've got a basketball doesn't mean we're going to make every shot because of variability, but we certainly can it's going to go through now if we go to the fairgrounds and they put that little tempting room right up in front of your face and give you a basketball so you can win this giant teddy bear by shooting the ball in that room you think it's easy um pay attention to the signal level because it's not a two to one ratio, right smaller room and perhaps he perhaps even a bigger ball uh but they're stacking the odds against you so that's essentially what's going on with six sigma so go back to the lean sigma game that we played earlier in round three when we went from that square diamond to the round diamond what did we do in terms of sigma put his six signal potential? We made the goalpost bigger right? So we actually made it easier to get the red circle in there so we designed in six sigma in fact, it's very difficult to get to six sigma it's almost impossible really, to get to six signal level of performance without designing it and you can't make your way there you've got, you've got to think about it from an engineering perspective and design it in so you design it in through clever engineering and design just related teo, you know, well, diane longer and is back in here saying she has a start up photography business is a solo preneurs can you just make it a little bit more of an example or or give us some concrete ways to actually take this sort of big concept of measuring and repeatability? Sure, and I don't mean it's not just, uh it photography they emitted anybody who's, listening online or anybody, anybody here, whatever business here and step number one to start to brainstorm, where do we see there's variation in our process to our customers experience variation in any way to a variation in our vendors variation and are different films films? Do we use? How many different photography instruments materials, machines to be his cameras? Whereas our variation in the business and if the answer is we don't have any you're not you're not getting it, so you gotta pay attention to where you have variation in your business in terms of product material performs go back to the issue cow a diagram that fish bone that we talked about were goingto look at that in a few minutes and think about ok from a machine from the material from a methods from a measurement systems where is the variation so just like we process map before there's variation of the process and think about it from a process standpoint so in my business what is the process that I used to get value to my customers map that process out and then search for the variation it's there it's there in a certain degree of variation is a good thing you certainly want just a little bit of variation going on with your heart is that true otherwise your flat line so I was kind of a joke but the idea is uh you know a little bit of variation is helpful okay but we want to be careful where it really hurts so that's just it and when we come back later to some some q and a and that I'd like to hear from people specific examples of we've got variation here we've got variation here we've got variation here and then we can start to root cause it and uh and get after it so this this this leads us to how predictable and reliable are we as a vendor as a supplier as a as a business owner um in our market another measure is six cpk cpk is the capability index just again another measure and you can read more about this stuff you don't have to get to too into the weeds on it these air simply measures to say how do we uh know what our signal level is? And you can find out by calculating some of these things or simply entering the data into a software like I said before and letting it do it for you that's the simplest way there's some great software's available to do that the voice of customer voice of process what are we doing? We're comparing the two all right, this is the car in the garage metaphor what what's the relationship all right, and what we know so far from our exercise in round one is that we're not going to get a twenty nine inch car into a twelve inch garage just that's not gonna happen. All right, so this would be a three signal process were essentially is a one to one ratio. So where did we have three sigmund little in the uh in the lean sigma game thick yellow circle department that yellow circle was almost the same with if you will or diameter as the ring wasn't it jane you were putting him in there it was a tight fit that's about a three sigma ratio you did remarkably well at a three sigma design okay, teo, get those in there without any heirs way didn't see any years there, but if we did it a million times, we'd see close to sixty seven thousand years, okay, in that kind of a design three sigma design. So by figuring out a way and we didn't do this in that round, we could have gone out and try to get get smaller yellow dots or get the ring's bigger, and we weren't we weren't allowed to do that in that round, but we did come up with a creative alternative in the diamond department to essentially do a design for six sigma uh bye bye thinking creative it creatively about that diamond shape. All right, that's, what six sigma would look more like? So all we've essentially done is going back here is we've grabbed this distribution and we've just pulled it up so that now we have a much, uh, a much better chance of scoring that's essentially six sigma, you know, in a nutshell, is we've we've we've gone after the, uh, the variation, and we've taken it out now, that's, what we're going to do with kaizen in the next segment is we're going to say, all right, let's, take that catapult let's, go back to our process map and let's start the brainstorm where the variation is and then where it's coming from using cause and effect analysis and using domestic and so I challenge the folks at home and online um take your take your business, take your process take take whatever it is you're doing and start looking for that variation where dr variation and billing and how long it takes to answer the phone and right first time quality and responding to customer enquiries it's it's amazing it's all over surrounded by you know I do a lot of times when I call it an eight hundred numbers customer service line and I need help on something and they can't help me I hang up and I dialed in same number again you know I do that because I'll get a different answer that's variation sometimes it works for you sometimes it works against you, but so often you hang up, you pick up the phone, you dial it back, you get a different person say here's my problem this is what I need and they'll help you where is the one the one before won't that's variation so sometimes we can actually use it to our advantage if we know those little tricks of the trade because not every customer service rep is following the same script or answering the questions the same way, okay? It's it's it's uh it's it's all over the place so just some measures again if if you're into this particular ah you know module or segment you want to learn more about six sigma these air just classic measures you can google this stuff you khun being it you can go after it and learn a lot more about it but the way I use this isn't a very user friendly way we don't need to get too into the weeds let's just let's understand what it is and let's focus more on the inputs that lead to these to these outputs alright so c p a to a c p k a one and a half and a defects per million is what that stands for sometimes you'll see dpm oh defect per million opportunities dpm adi pmo of three point for that street point for defects per million opportunities would be sixty six sigma level of performance so we go from defined out and measure all right, we've defined our current state, you know we've gone ahead and mapped it out appear and and uh that's not shoot it is alright I mapped it out we've done the process flow we could pull out the aipo diagram we could start looking at other tools like the history graham the run chart the control chart to get a good idea just what is our process currently capable of someone walking through a few of those now here's an I p o diagram is an illustration of how an I p o diagram can be blown out of the like of manufacturing example all right so if in this case we've got raw materials we've got manufacturing packaging distribution customer and we say well that's we want to zero in on manufacturing and run on I po on that what would be some and it happens to be tablet manufacturing what would be some of the inputs that are critical here in the outputs that are critical to measure here so starting with the outputs we want to measure the tablet wait could we have variation with the weight most certainly thickness we'll certainly hardness potency dissolution ruthie's air critical measures and this is a real example so all right what? Um what are the critical inputs over here while the amount of ingredient a the amount of the ingredient b the pressure that we compress it at things like that time temperature, humidity those air all this is this is where we have an opportunity for variation to enter the equation and this is where we're going to experience it so variations always propagating from this side to this side and so the secret being we've got to come back here and we got to control this if you ever see a really good recipe it's very precise you know this amount of this and this amount of this you put it in the oven for this amount of time, ok? If you follow this recipe very specifically, you will get predictable outputs wait, actually, variations that are being shared ready flame ire, the different names on tv the other day there variations are client outcomes and client expectations result of lead generation and other types of marketing. So would that be spot? Sure. So the result is we've got that variation in the with the client, the customer, and it sounds like the second half of that was we've already gone upstream and figured out at least one of the reasons why and have the lead generations there's lead generations, perhaps another marketing. Okay, so what we're doing now is we're capturing the variation in our process and were captured. We're showing it over here so this could be the distance the ball goes. It could be the amount of time it takes to generate a lead. It could make it be in the amount of time it takes to fill up on order. Okay, it's. Everywhere we captured over here. And then we use a tool like the aipo to gain knowledge about it's variation over here. And if it's undesirable, then let's, take it over here and figure out what's causing it. Maybe the oven's turned up too high. Maybe we've left it in the oven too long. Maybe we didn't put enough sugar in the rest of peace, so speak metaphorically, but there's going to be something going on over here that's causing that variation and lee generations and and, uh, inclined experience, this is so this is this is how we'll use it. I'll come back and try to continue to build and build on that with the examples that we have variations that are outside of your control. Oh, yeah, yeah, I'm gonna cover that we call those noise variables so there's, controllable variables and there's noise variables. So, like in the airline industry, look, we can we can control a lot of things, but we can't control the weather. So if we're going to fly and hot and cold and rain and snow, we've got to make a process real bust is the typical term to that we've got to make it insensitive to those noise variables, or we're at risk, and that actually can relate back even to photography and baking in like a weather delay. If you're doing a commercial shoot like that's, right, this is where failure mon effect analysis comes in, where we say, what could go wrong? Look, a rain so what's our countermeasure what's our what's, our pre programmed response we're not gonna wait until it's raining to figure out what to do. We're going to have a pre programmed response a back up plan. All right, that's, what smart companies do they have backup plans? Not like they're caught by surprise when it rains because it's sooner or later it's gonna it's gonna rain you know what I'm saying? Okay, so now we could actually dig down even deeper let's just pretend in this example again, this is just an illustration that we've got problems going on with our dissolution rate. The tablets aren't dissolving the way they should. Customers are complaining again, you can use any example you want here so we could say, well, let's drill down deeper into dissolution rate might be there now, it's just the distance that's what we're really focusing on here what's going on with that? Well, dissolution is really relies on these x variables, you know that you want to die the amount. Tell him ana lak toasty amount of wax amount of magnesium but potassium this is a real example, by the way. Um, so we need to again investigate more to figure out just exactly what that cause and effect relationship is if we're going to fix our dissolution right so this is again it's a knowledge gaining tool so we start with the variation wherever it is that's on an output and then we go from there everybody should not be thinking about where do I see the variation? And then we can display the variation, display the data in different ways so one way would be a history. Graham okay, so we've got now we've got a number of shots landing between two and four a number of shots between four and six number shots between six and eight, so to speak, things like that and you can actually see now if we threw a blanket over this we'd have that normal distribution that curve that we want to then used to statistically learn more about our process and there's information now in your material is well on how to actually create a history. Graham the simplest way to do it today is put the data into a spreadsheet and click on history just walk you through all the mathematics behind it uh but it's here for your leisure. All right, we run shirts I love run charts because you could just look at and go really that's what our process looks like, you know, you could just see the variation in a different kind of way, so if this is the amount of time it takes to fill an order so we put time up here, and it takes us on some days it takes us four hours to fill in order and other days we can fill it instantaneously, and on other days, we take eight hours or something like that to fill an order. Okay, we can actually if you just think about it, you could throw a great big distribution curve over this because we could calculate an average we could calculate a standard deviation. You turned that ninety degrees, you've got your your normal distribution. This just visually shows us. And then every time, this is like, isaac example of golf, you know, you finished the golf hole? What do you do it right down your score? So we finished our changeover, let's write down how long it took. We finished doing a processing our film, we finished. Uh, how many? How many deals did we sell this week or this month or today? But you can use this on anything. Write down your score and then you could start to really learn about your performance from your data. A lot of times I go into companies, they say, well, show me the day that we don't have it, we don't keep sort we don't keep that score, we don't have that data case well, if it's important that we should probably start so let's start the start measuring and keeping track of what's important let's get that done and then we can how we supposed to gain information and intelligence and knowledge without data? So data's data collection is a very important step in all of this from the run charts we can build control charts and again, the easiest way to do this is load software up with the, uh the data points and said I wanna control chart is going to ask you what kind of control charges different types of control charts and expert on our charts and yu charts and p charts and things like that. So you got a research a little bit into just what kind of chart you want to use it, if you will, but this is going to do just what I just showed you from that data we get enough data points uh, we could calculate the average the mean we can calculate the standard deviation we can populate the upper and lower control limits, okay? And we've got nowt we've now got a model to use um, to monitor, to measure and to use to improve our process, whatever that happens to be and it's we're talking about minutes a day, not hours and hours and hours it doesn't take you long to write down your golf score unless you play golf like I do and you can't find the ball all right so again here's ah little information in your booklets on control charts and how to do those again the simplest thing to do is let the software help you a little bit the bottom line here those we want to make sure that we can determine the stability of our process how stable I'ii how predictable is our process you look at that catapult that we shot earlier how predictable and stable is our process you'd say it's not we don't know where we don't know where the ball's kind of land next truly with any degree of certainty other than it's gonna land probably somewhere within that twenty nine inch range but if I said I want you to shoot it into a cup where do I put the cup you want me shoot the bond with cup like a little coffee cup yeah that's exactly what I want that's our six sigma test for the next session where I put the cup I don't know all right so in the next round we should know we're gonna put the cup right here and we're going to see if we can shoot the ball in the cup then statistical process control it's just a way of listening to our process so now we can listen to it we can monitor and from there we can begin to interrogate it a little bit and we could start to read for trends. So what happens, for example, if we've got a whole lot of data points on one side of the center line, what's that all about that's like flipping? In fact, the magic numbers statistically is seven. If we get seven data points in a row on one side of the means or the other that tells us our process is out of control, the reason for that is puts the odds of flipping a heads or tails seven times in a row with a coin because to do it once it's fifty fifty right to do two heads in a row is point five times is point five point two five about three heads in row four heads, five heads in room when you get to seven it's less than one percent point oo seven so now all of a sudden we just flip seven heads in a row or seven tails in a row that tells us something fishy is going on with our process. We need to investigate it. So we use simple statistical measures to monitor a process. Okay, to find out just how stable it isn't, how in control it is, and if if the data tells us there's something fishy going on, it tells us to investigate, and by the way, let's investigate before it goes hey, wire goes out of out of control. We have a lot of cost of poor quality, but we have two different types of things going on here. One is a place called special cause. And if it's a special cause situation um, our strategy is number one. Get timely data. Get something immediate. Er, we've got to take special immediate action on it. Some something something crazy went on. Okay, we're out of bounds. Okay? The space shuttle blew up. Something serious happened? That's uh, that's not supposed to happen. That's not normal let's let's. Check it out and and learn from it retained the lesson. So that's what we classed by his special cause? Spend more time on this and the next segment is, well, common causes normal variation. This is its normal supposed to happen so we've got some normal variation going on all right, but let's not try to explain away every little data point on a run chart every little one it shouldn't be flatlined flatline would being dead potentially. So we're just trying to treat a common cause like a special cause that we're actually going to make the process work let's just let's understand the difference, alright or otherwise, we could actually make things, make things a whole lot worse so we come back to this food for thought, our systems and structure perfectly designed for the results we're currently getting. So right now, our system from round one is perfectly designed to get us the twenty nine inch range, okay, so if we're before good with that, we could go on and kaizen something else, because we got data to prove that we're good at getting twenty nine inch range. On the other hand, if we learn from the customer where that's just not acceptable, I got to see a six inch range, okay, not a twelve inch range, but certainly not the twenty nine inch range. We got work to do. So that's, what we're going to do in our next, uh, our next segment is we're going because we're going to get into, uh, how do we drive out that variation?