Save Teller Initials with airSlate SignNow

Save teller initials

this video will step you through developing a simple model using the microsites sharp discrete simulation tool microsites sharp is a commercially available general-purpose discrete event simulation software tool that was first introduced commercially in 1984 since then the tool has undergone many changes and enhancements micro st. sharps intuitive graphical user interface and flow chart approach to modeling make it a tool that can be used by generalist as well as simulation experts my Christine is proven to be an invaluable asset in both small businesses and fortune 500 companies and in many areas including manufacturing the military human factors healthcare and the service industry do you want to know where the name came from Saint stands for systems analysis of integrated networks of tasks the micro part came from an early contract to redevelop st. as a micro computer tool the sharp portion came from the tool redesigned to use the c-sharp language the tool has a long history of being in the forefront of computer simulation microsites sharp is a robust full-featured tool this slide show just a parcel list of some of its features such as 2d and 3d animation optimization input/output communication capabilities free runtime export it can compete or outperform any tool on the market what I'm going to illustrate today is how to develop a simple model of a bank the simulation question is how many bank tellers do we need to ensure that customers do not have to wait too long while ensuring that the tellers are utilized now let's go to my croissant sharp this is what you see when you first open my croissant sharp you will notice it's a fairly standard Windows application by default there is a menu bar a toolbar treeview a message bar properties window descriptor window area and output window the environment is user configurable you can move resize hide and unhide various components that you desire the start page is by default shown in the middle we want to start a new analysis so I will just click the new icon up here on the toolbar notice the start page closed I have a blank slate for defining my model I will now create two tasks by right-clicking in the network and selecting one of the options in the menu the first task I will create it will represent customers arriving so I would select specialized tasks of add entity generator so it can generate those customers the second test will represent the customer being served so I'll add a regular task to link the tasks together to move the tasks I just grab it and move it to draw a path I started the first task my cursor changes to a pointer and I simply draw a line to the second one now I have my two node model now I need to add task information to open the task description window just double-click on any any node so I'm going to double-click on the first node here now I'm going to enter a name the name of this one is customer arrives notice there are four tabs below the name main timing paths and appearance and notes this is a specialized entity generator tasks will generate entities based on the rules supplied by the user in the main tab in this case we want this task to generate customers for an eight-hour day or from time 0 to time 480 minutes so I'm going to change the repeat type here from run to count to run to time and then I'm going to enter 480 in the repeat duration field notice that by default the code entity tag plus plus is in the code field this ensures that each customer generated is unique and has a unique tag value so I'm going to leave that as is now we would go to the timing tab as the name suggests this is where you will set how often the task occurs for a typical Bank the arrival rate of customers will vary depending on the time of day and during lunch the arrival rate was increase so I'm going to use a variable named arrival rate as the mean time value for this task so I'm going to change where it says return zero to return arrival rate keeping the semicolon there then I'm gonna select that word oops I typed it wrong so I'm gonna select that word right-click and I can go down here and say add selected to variable list and I'm going to add it as a floating point so it's going to represent time now it's going to take that word and enter it over here and to my tree view under the variables so it's now added to the variable list now I'm going to also change the distribution from normal to exponential just by picking it from this drop list here the next tab over here is paths and this shows me that my path I have one path and it's gonna go from this task to task two I can close a task description window by just clicking the X up in the upper window now I've got my first task and it kind of overlaps my second one so I'm just gonna grab it and move it over here so now I've got my first task to find let's open task two by double clicking on it and we'll enter the name serve customer we only want this task to execute if a teller is available so we're going to use our release condition and in this code block here is where we enter code that says what has to be true for this task to execute in our case a teller has to be available so we're gonna create a new variable called tellers and we're gonna say that you have to have one at least you have to have a greater than zero for this task to execute or this release condition says return tellers greater than zero then the beginning of text says what happens when this task starts so in our case we're going to decrement the number of tellers available whenever this task starts and then once the task ends when we're done serving the customer we're gonna make that teller available for the next customer so we're gonna say tellers plus plus semicolon so now here in our release condition it says the tellers has to be greater than zero at the beginning effect we're going to decrement the number available and at the end effect we're going to make more available and we need to create that variable so I'm gonna go variables and say add variable I'm gonna give the name of tellers I'm gonna say okay it's gonna add it down here now notice the property grid here is updated with that and the initial value is initially set to zero I'm gonna make that two so now we've got tellers to so at the beginning of our model it's gonna come in here test this release condition it's - so is that greater than zero yes it is let the TAS release go to the beginning effect decrement it by one we'll have one teller and after this task is complete we incremented again and we'll have two left so now we've done our coding for releasing with the tellers now we're going to go to the timing tab and this task is gonna have a gamma distribution with a mean time of six so I'm going to put six in there and a standard deviation of two now we're gonna go to the Q tab in the bank customers are served in a first-in first-out fashion so I'm gonna change this Q type from none to FIFO first-in first-out now I can go ahead and close this task as you can see I now have a customer arrives followed by serve customers and my serve customers has a queue in front which is a FIFO as indicated by the F in the square in the front of the box now we need to setup the customer rivalries we want an initial rate in the morning a faster lunch rate and then a rate after the lunch rush in my croissant we use events that are called scenario events these happen at specific times in the model to add a new event I'll go over to my tree view right click on a scenario event say add scenario event notice a new one gets added there I'm going to double click on it to open it in this first one I'm going to call initialize it's gonna happen it's time zero and I'm gonna set my arrival rate equal to three okay so I'm going to close that and I'm gonna add another scenario event double click on that one I'm gonna have this one happen I'm gonna call this one lunch lunch rush and I'm gonna have that happen at time 180 which would be three hours after opening and then I'm gonna set arrival rate equal to two I'm gonna have a customer come in about every two minutes okay and then I need to reset it get after lunch so I'm gonna add yet another one this one's gonna say rush over I'm gonna set the arrival and that's gonna be at time 240 and I'm gonna set the arrival rate back to 3 go ahead and close that so now I've got three different scenario events initialize happens at time zero lunch rush happens at time 180 and rush over happens at time to 40 and it resets the or customer arrival rate I now have a simple to note model with different customer arrival rates let's save the model as Bank st. okay now before we run it let's check the syntax so we go up here and click the little check mark on the toolbar and down here in the output window it says the code generation was complete there's no hours errors found and it was created successfully if there was an error it would be displayed here let's for example let's open up our serve customer let's say I added another code code here with the new variable but I didn't define that into my variable list now if I go through if I'm gonna close that and I check my syntax again it's gonna come down and it's gonna tell me that there's an error there there was a new variable and it expected a semicolon and also doesn't know what it is so now if I click on that error it's gonna take me right where it wasn't underlying it so it tells me where the error was and then I can go right to it okay now I've deleted that let's check it again okay no errors found so let's go ahead and run the model I just click the begin simulation tool on the toolbar okay that Ram pretty darn quick it went to 485 and took 7.5 real-time seconds so now let's try and slow it down a little bit we're gonna single step so either I can click the single step up here or I can hit ctrl T and then as I go I can see the task Network update I can see customers arriving I can see them being served I'm just keeping hanging ctrl T here and then notice down here in the corner is the clock value so I can watch the clock as it goes so we're now at time thirty minutes so bank's been open for about a half hour now my cue starting to build up I've now got sweet people on my queue and two people getting served so they keep going I mean I've got five people in my queue so that means I got pie people waiting in line so if I keep kind of stepping through here I can see I'm you know kind of about five I hope I had six people there and I think it's about 180 is when we start getting even more so now my queue starting to build up I've got nine people ten eleven so I can notice that my queue starting to get pretty big now I haven't even gotten to the lunch rush yet so at time one eighty now I'm into my lunch rush I've now got fifteen people there so I'm gonna go ahead and finish running and I'm just gonna hit ctrl G to finish it so so I can see that my cue was built up just by watching that we already know by what just watching the model but like you got pretty long let's look at the reports and see what they tell us in my croissant the reports are stored under something called snapshots so if I go back to my tree view and I expand that snapshots node I can see the built-in reports let's look at the queue data collection so to look at that I'm just gonna right-click it and then say show result data and that's going to open my queue so I can come in here and I can see my maximum queue length was about 19th and the maximum queue weight at this time that was the mean was 28 the maximum was 58.5 so somebody had to wait in line almost a full hour that seems a little bit long I don't think my bank customers would be too happy about that so I might close that now let's go back and change our tellers and now instead of having two tellers let's make something like five so I'm gonna go ahead and save that I'll run the model again okay it's already done so let's go back and look at our queue data collection this time this time let's see our maximum queue length was three and our maximum queue weight was only two point four so that's not bad I think you know our customers wouldn't wouldn't mind meeting two and a half minutes so now I can either keep going back and changing the number of tellers and looking at that result or I can use it use some built-in functionality in my Christine that helps us I can use either the experimenter down here or I can add use the OP request optimizer let's go ahead and use the experimenter so to add a new experiment I right click say add an experiment and I adds a new one below there let's open that up so our first experiment we're gonna add our input is gonna be tellers so we'll go pick that we have to have a minimum of one because if we had zero we're gonna have people waiting forever let's have a maximum of I'm no 10 and with a step of one then we can tell how many times we want to replicate this experiment let's go ahead and set it up to five so that means it's gonna run it five times for each input variable and then what do we want for output for outputs we want to know how long the simulation runs so let's pick clock so pick clock there so now I have a new experiment I'm gonna save my model so if I go over here and I say under experiment menu I say run it's gonna go ahead and say ask me which one I'm gonna run it Micra st. quickly went through and ran through all possibilities of tellers and ran each one five times so if I look if I only have one teller here it ran five times and it took anywhere from 928 minutes to about 1191 so that took a really long time if my bank bake was only open four hundred four hundred and eighty minutes and it took twice that long to get through all the customers so I know one isn't very good let's see let's look at to choose a little bit better my bank closed at 480 I'm one of these it got done at 485 so that's not too bad so get number three here so those are actually a little bit less my three tellers are pretty good I get to four tellers everybody's done I was able to close a bank within ten minutes after it closed but after this once I get past four here I'm not getting much return on my investments so adding more tellers isn't giving me much time so that's telling me that I probably only need somewhere between three and four tellers and I'm not gonna have too big of a queue so there we go we've got a simple model with a bank with customers arriving and serving with different weights during the day and we've used experimenter to look at all the possible values of some of our input and output variables thank you for taking the time to watch this video and I wish you happy modelling you can download a demo version from WWE microsites sharp comm or send any questions to my christine sharp at alliance science com