HeavyBid Monthly Webinar May 2024

Alright. Welcome to the estimating monthly webinar series for the month of May. Bear with me just a moment while, everybody finishes climbing on in, and then we will get started. Alright. It looks like we're starting to level out as far as participants are concerned. So we're gonna be, we're gonna go ahead and get things kicked off here. Welcome to the estimating monthly webinar series for the month of May. My name is Aaron Allshaus. Guest speaker today is doctor Sam Savage. And then also joining us to help facilitate questions and answers is Paul Lambert, the technical product manager for preconstruction. What we're gonna be talking about today is risk in construction. So what is risk in construction? Basically, risk is uncertainty. It's fear, it's uncertainty, and it's doubt in what you're doing. So to give everybody a little bit of a background into where this conversation started, it started probably about ten to twelve years ago here at HCSS with how do we account for how do we account for the unknown? And it was something that we kinda toyed around with, for four or five years until we had a customer come into the office, one day. And just by sheer luck, I was in the room with them. And we were kinda going over just some new features that have been in the program, for the last year or so. And they started talking about how they how they, account for risk. Well, Mike Ryden, the the founder of HCSS, was in the room as well. And if you know Mike Ryden, you know you know his his ears perked up, his eyes lit up, and for lack of better words, the the rest is history. So, basically, after that conversation, we got we got to work on developing a risk module in heavy bid, and we're gonna be talking about that today. But we're also going to be talking about risk in general and talking about how we account for that. Now without further ado, let me introduce you to our our guest speaker, doctor Sam Savage. Doctor Savage is the, professor of environmental engineering at Stanford University. He's also the the director of probability management dot org and the author of the flaw of averages. Basically, he has he has lived, ate, slept, and breathe, probability management and risk, for the last several years. And he was gracious enough to to offer his time to come and share his thoughts and his, his feelings on this, this topic that is extremely exciting for him. So, doctor Savage, I'm gonna turn it over to you, and we can, we can get started here. Hey. Thanks a lot, Erin. If if you will then stop sharing the screen and spotlight spotlight my screen because, I'm gonna be using my my green screen here. So pause if you can here's I I maybe can do it for myself. How's that? Okay. So, let's start with, this concept that I call the flaw of averages. And, basically, you may have heard of the statistician who drowns in the river. It's on average three feet deep, kind of a classic example. What the flaw of averages states is that plans based on average assumptions are wrong on average. Now we have to ask ourselves how this tragedy played out. And what happens is the boss says, how deep is the river? And you say, it varies, boss. And the boss says, give me a number. Has anyone heard this phrase, give me a number? So we're gonna get a show of I'll bet you have. Anyway, that is a fork in the road to hell. Now the flaw of average explains why so many projects, are behind schedule, beyond budget, and, below projection. And let's begin with, oh, just one sec, with a with a construction example here. So we've got a a set of permits that have to come in before we can begin construction, water, storm drain, you know, sewer. It's not our first rodeo, and and we know from past experience that on average, each permit takes six weeks. But, of course, it could take eight weeks for water, say. And there's really only one formula in the worksheet here. There's a max statement down here. So you obviously can't finish the project. You can't start construction till the last permit is in. Now one other thing to notice, there's a penalty clause. So if if we don't finish if we don't start construction within seven weeks, we're gonna incur a hundred thousand dollar penalty per week. Now let's take this thing back to six weeks. And, you know, the the boss comes in and says, when do we start construction? And, you say, I don't know, boss. I don't know how long water, storm drain, blah blah blah takes. Boss, of course, says, give me a number. And you say, well, you know, each of these takes roughly six weeks, so I would guess we'll be done about six weeks. But you can't hold me to that because there's a lot of uncertainty. So now we're gonna have our first poll here. What is the chance let's let's remember. Let's go back to this thing. Right? These can all vary, but on average, they'll be six weeks. So what's the chance the construction will start within six weeks? Hundred percent, fifty percent, ten percent, one in a hundred, one in a thousand, one in ten thousand. If we can start the poll, let's just see what people think. By the way, people in the construction industry are pretty good at this compared to people in other industries because they've had to really work on these projects. So do we have the poll going here? The poll should have it up momentarily. Yeah. We're getting ahead of it right here. Just a moment. Right. Yeah. I'll take I'll take you back to the spreadsheet. Right? So, you know, that one could be five weeks. That one could be seven weeks. You don't know what they're gonna be, except you do know on average, they're gonna be six weeks. Alright. Savage, do you mind switching back to the, other screen one more time with the choices? So how are we doing on the poll? And I don't want I don't want others to see the poll results because I don't want them to be influenced unduly by what the other folks are saying. I can tell you when we're done, I'll tell you what what most people think or what a lot of people think. We give it up right now, so everybody should be able to, participate that. Very nice. Okay. This is instructive. So I'm seeing it as the host. Is everybody else seeing it or not? Yeah. We have, all thirty seven people have answered. Okay. Excellent. Okay. So now okay. So now we can end the poll. I want, I wanna share the results, and let's take a look. Okay. Can everybody see this? So the winner is ten percent. Often the winner is fifty percent. The right answer is one in a thousand. However, I mean, would that make a difference? Oh, here are two airplanes. This airplane has one chance in ten of killing you in a crash. This one has one chance in a thousand. Which airplane you're gonna get on? Of course, it makes a difference. Okay. So let's stop sharing this. Folks, the right answer is one in a thousand roughly. Okay? Oops. Stop sharing. Oops. Can I let can we close that window? So now the question is, some people got it right. Maybe they just guessed. But how do you explain this to the boss? And and let me do that for you over here. Okay? Look. I didn't tell you this, but just suppose that for each of these, permits, there's like a fifty fifty chance that it's less than six weeks or greater than six weeks. Then the only way you start construction in six weeks is again, all ten of them are less than six weeks, but this is like a coin toss. So the chance of finishing within of starting within six weeks is like flipping ten heads in a row, which is like impossible? No. There's about one chance in a thousand. Okay? So that's how that works. And, Aaron, if you wanna, take over here, and discuss what, what HeavyBid has been doing in this in this regard. Interestingly, when we met up, I realized that HeavyBid has been putting in a lot of the infrastructure to really do this right. And this has been an exciting project working on, on this with them. Oh, right. So jumping back into heavy bid. As I alluded to earlier, we we had a lot of discussion internally. We had some discussion, with external customers on how we want to track risk. And, basically, what we what we decided on was in heavy bid. If you're on a comprehensive system, within the activity information, there is gonna be a tab called risk. When you go ahead and click on that tab, we thought there was three different types of risk that essentially customers want to account for, and that's productivity rate range. So if I have a crew and I have a production rate on any given activity, I can give a best case, worst case scenario on my production rate and track my risk that way. I have a cost range to where the cost for my activity is one numbers. In this case, sixteen thousand one hundred and change, and I can give a best case and a worst case on my cost as well as a cost factor change. In this case, it's going to be, my my best case is on a factor of one point one or point nine point nine times whatever my cost is. Those were the three different risk types that we thought would be appropriate, for the end user. So for the sake of demonstration, I'm gonna go ahead and I'm gonna select productivity rate range. And my productivity, my production rate for this activity is fifteen hundred units per shift. In this case, I'm doing fifteen hundred linear feet per shift. But let's say for sake of demonstration, my best case is I can do, I can do eighteen hundred units per shift. And the chances of doing those eighteen hundred, let's say, are ten percent. But, conversely, my worst case scenario, let's say everything goes wrong, I'm only gonna do eight hundred units per shift. And you know what? I've I've got a I got something in my stomach that's telling me that we might have a better chance of hitting that worst case. I'm gonna give it a fifteen percent probability of hitting that, worst case scenario. Basically, I'm gonna go down and I'm going to select all the items in my estimate that have that are risky item. Essentially, anything that I feel could have a best case or a worst case scenario on any of those items. Now there's another option in here called include as risk contingency. Basically, what that does is it's gonna say this activity is specifically in my estimate to account for risk. Now down at the very bottom of my estimate, I went ahead and I created created a bid item, called contingency, and this is an indirect bid item, just for the sake of demonstration that I've also included as a risk contingency. Now I don't need to put in any kind of particular factors or anything like that. I just need to make sure that the dollars for that activity are in the estimate. In this case, I've got fifteen hundred dollars that I have set aside in my estimate to account for my risk. Now how do we analyze all of this information? I can go through and I can put all this in, but it doesn't do me any good if it doesn't basically tell me how much money I need to account for my risk. So within my subsystems menu, I have a button called risk analysis. That's going to take me to a sub menu here where I have my four different categories of risk. In this case, productivity risk. You'll see the activity we created as a productivity item, and it's giving me my, my actual production rate and it's giving me my best production rate and my worst production rate. If I want to modify those, I can do that from here as well. I can make the modification here without having to go back into my activity and make that change. I can then go over to my other productivity risk or my other activity risk. In this case, I have my risk contingency sitting right here. The reason it's showing up under other activity risk is because it is not a productivity risk. It's my, it's my cost range or my cost factor range is gonna show up here. I also have the ability to account for subcontractor risk. Now the reason I didn't show this in the tree view is because this is only done here within within the, risk analysis module because it's built based on what's in your quote system. So in this case, I have a subcontract folder in my estimate that has a vendor total of twenty thousand dollars and change. Now why would I have a subcontract risk item in here? The reason for that is the probability of rework. What are my chances that this subcontractor is gonna have to do all or part of their work over again? So I can put in what are the chances of having to do rework as well as what percentage of that of the item that the subcontractor is quoting is going to have to be rework. So the chance of rework plus how much of it is gonna have to be rework. Now I also have the ability to put in, add ons or use my add ons as, risk contingency. So if I have any overheads or instead of trying to create an indirect item like I have, I can use my add ons as an easy way to put in a lump sum value to account for my risk. Now what we're gonna do is we're gonna come down. We're gonna click analyze risk. What HeavyBid's doing in the background is it's running what's called a Monte Carlo simulation, which is basically saying it's running through the scenarios ten thousand times over, and doctor Savage is gonna talk a little bit more about this, shortly. But, basically, it's gonna go through those scenarios, and it's gonna print out a report telling you how much money you need to have in your estimate to account for your risk in order to meet, certain variable thresholds. And what I mean by that is as soon as my report finishes loading, I have a report here that's basically giving me a risk summary. In this case, I have my total project cost. How much money do I have in risk contingency, which gives me my base project plus my risk, which gives me my project total here. I have my markup and I have my bid total. Now we give you three different risk scenarios. Basically, out of those ten thousand scenarios, if I want to account for fifty percent of them, in this case, I can actually take just shy of five hundred dollars out of my estimate if I wanna account for fifty percent of my scenarios. If I wanna account for eighty percent of my scenarios, I need to add an additional nine hundred dollars to account for the to be able to hit that eighty percent threshold. And if I wanna hit the ninety percent, which basically, again, is just saying that of the ten thousand scenarios, nine hundred of those scenarios, I have enough money in my estimate to account for them. I need to add an additional one thousand six hundred and twenty nine dollars. Now if I wanna dig a little deeper, if I have a lot of items in here, I can go through and I can see all of my all of my breakdown here. So in this case, like, for subcontractor risk, I've got my potential rework cost here to give me my potential or projected total cost. So, basically, it's just my estimated cost plus my potential rework cost. So this is how much money I could potentially have to have in my estimate to account for all of the work that subcontractor is going to do, potentially. Again, they might not have any rework, but we always wanna be safe. And, again, I can see my erosion control item, which is where I put that, that clear and grub item in. Now that is essentially what we have in heavy bid right now where you can go and you can track based on, production rate, cost, or cost factor and track how how much risky dollar you how many risky dollars you have in your estimate as well as how much money you have in your estimate to account for those, those risky items. With that, I'm gonna turn it back over to doctor Savage, and he's gonna, talk about, a really exciting new feature that he's been working on. Hey, Aaron. Thanks. Thanks a lot. And, I wanna say you have now seen the infrastructure in place, to do a lot of this stuff. And here's how the world has changed. So I wanna talk about the chance age. I'm gonna start with the simulation age. And Aaron mentioned Monte Carlo simulation. Alright? I'd be interested. I don't know. Paul, can you check show of hands how many people have heard of it? But let me give you my definition of it. What's the last thing you do before you climb on a ladder to paint the side of your house? You shake the ladder to see if it's stable. That's doing a Monte Carlo simulation of the ladder. You're bombarding it with random forces to see what it does. K? I'll bet I'll bet everyone's done that. And I have to hasten to point out, I'm sure everyone has heard the term garbage in, garbage out. It does not apply to simulation. Let me tell you. The forces on a ladder when you shake it are not the same as when you climb on it. So how many people gonna stop shaking ladders because I told you you've been using garbage all your life for the estimates of the forces? No. You're gonna keep shaking the ladder. Of course. Oh, Monte Carlo simulation. Be interested. Oh, quite a few. Great. Great. And now you know how to explain it, to your boss. It's like shaking a ladder. Okay. Let's let's that that's that's great to see the percentage who are who are interested. And that, by the way, is exactly, what Aaron demonstrated for you. Classic proven approach. And, so now, let's see. Can we take that off the screen so I can, continue here? That's great. So more than half have heard of this, and it is the underlying technology, and it's been around, since nineteen forty seven with the Manhattan Project. The way to think about it is this, that, that you you, estimate chances by generating random number on a computer and counting the number of times the events of interest occurred. But I now take you to the chance age. In the chance age, the chances are embedded in data. The data element is called SIPs, stochastic information packets. But but let me give you an analogy of this first. Imagine that instead of illuminating the chances of something, you wanted to illuminate your book so you could read it at night. Now starting around eighteen ninety, you could have gone down to Thomas Edison's light bulb laboratory, and Thomas Edison said, hey. We got a generator in the basement. We're cranking on electricity. Come on down to my light bulb lab with your book, and you can read your book at night. And then a bunch of stuff and that was the power plant and the light bulb. Then a bunch of stuff happened where you don't have to go down to the power plant. Using the sixty cycle AC current standard, you can run the power out of the power plant to your bedroom, and you can read your book in bed. That's what's going on here. So HeavyBid already has a power plant in it, and I run this nonprofit that has basically developed a way to distribute this uncertainty, over data. So first of all, I have to define for you how you behave in the chance age. It it's a different mode here. In the chance age, when the boss says give me a number, you say, what do you want it to be? Here are your chances. And how do we do that? Well, I'm gonna show you now. We'll start with dice. I love dice. They display a lot of uncertainty. Right? And and by the way, an oil, economist once said told me years ago, he said, wouldn't it be fun to make average dice? I said, what would those look like? Oh, well, they'll have three and a half dots on each side. They'll be flat. Right? And you flip them and you always get three and a half. Well, no one would do anything that dumb. Yeah. As soon as you, you know, give the boss an average, you're basically playing with flat dice. Let's look at real dice. And one of the magic things that happened here is that about ten years ago, native Excel became powerful enough to roll dice or anything else for that matter. So how do we get uncertainty data? These are three columns of ten thousand rolls of the die, of each die. They're called SIPs. Right? And there's roll one, a four, a six, and a five. Role nine thousand seven hundred sixty five is three ones. I seem pretty confident of that. There they are. Why is that important? Well, accountants go nuts when they hear you're doing Monte Carlo simulations. Oh, you're generating random numbers. With the discipline of probability management, which means storing the uncertainty as data, When the accountants come by, you say, oh, no. No. This is a, this is a deterministic model, and there are thirty thousand numbers here. You guys are good at auditing stuff. You audit these numbers and come back when you're done. Don't miss a single one. Okay. You just got rid of the accountants. Now let's go further here. I am going to roll the first die in native Excel. And by the way, for you Excel gurus, it uses something called the data table. So there are no macros or add ins here. Alright? So I click on that, get out your stopwatch on your mark that said go. Did you time that? No. You didn't. It was instantaneous. That's important. It's interactive. And what happens is, oh, yeah. Equal likelihood from one to six when you roll a die. What happens when you roll two dice and add them together? Some of you may have played Monopoly or Craps, and you know that the most likely number is seven. Why? Because there are more ways to get a seven. Right? A two and a five, a one and a six. There are a whole bunch of ways to get a seven, and only one way to get a two, two ones. One way to get a twelve, two sixes. Okay. But we're in the chancing. Oh, so now let's do one that you you don't know the shape of. I'm gonna multiply, two dice together. Holy Toledo. What's going on there? Anyway, we're in the chance age and the boss comes in and says, I understand that our our company is multiplying two dice together. What number is going to come up? And you say to the boss, what do you want it to be? I'll give you the chances. And the boss says, well, I want the number to be greater than eight. And you say fifty five point six percent. How do you do that? Nothing to it, folks. It's just a count if statement. It's just a count if statement. So we count up all the times the result was greater than eight, and we divide it by ten thousand, which is the number of trials. And by the way, in our Excel version, you see, we've always got ten thousand results over here. So now you've seen dice, in the chance age. Get back. Get back here. And and let's move on now to the when do we start construction problem. So the first thing to say is that we've replaced the, permitting times, actual permitting times, basically, with die rolls that simulate permitting times. Okay? How would we do that? Well, why did God give us statisticians? There are a few people who know how to do this, and we're taking their word for it. You you don't ask how people are generating electricity that you're using in your light bulb tonight. The thing that's storing the uncertainty in data does is like a division of labor. You got a tiny fraction of people who do understand the stuff, and like the rest of us, I don't I wouldn't know how to do this. I have some ideas, but you go to experts for that. But look at this. We're not in Kansas anymore. Okay? Where are we? Well, I don't know how many have heard of Rick and Morty, but we're in a Rick and Morty cartoon. Rick and Morty is a cartoon series about these parallel universes. So we're in, like, ten thousand parallel universes here, and, I I can always go back and look at the average case if I want. And by the way, the maximum of ten sixes is six. Right. But the boss says, when do we start construction? And you say, when do you wanna start, boss? And the boss says, well, remember, the chance of starting in six weeks is, like, virtually impossible. But the boss says, well, look. We get penalized after seven weeks, so I certainly don't wanna start after seven weeks. And what's happened is when Aaron was running, the heavy bed, the simulator and it went chuggedy chugged, it's shaking the ladder a whole bunch of times. In our case, the ladder is pre shaken. And if I look at this, it says, oh, boss, the chance of finishing in less than seven weeks is four percent. How do you like them apples? Boss says, I don't. How about finishing within eight weeks? Twenty seven percent. Oh my gosh. Another thing we're gonna ask is, on average, when will we start construction? That is, what if I had a thousand or ten thousand of these jobs, and they all have the same uncertainty in permitting times? So given the average assumptions, we're done in six weeks. On average, nine weeks? Yikes. If you're planning for six and you know that on average, it's nine, that is not close enough even for the construction industry. Now let's go back over here. Let's look at our penalty. Oh, let's assume all the average conditions. Oh, there is no penalty. That's great. Yeah. But what would happen on average if you repeated this in a thousand, you know, Rick and Morty cartoons or doctor Strange in the multiverse of madness, think parallel universes. Oh, no. On average, two hundred and three thousand. You know, Aaron was talking about contingencies. Yeah. Yeah. You bet you better I show this to you at first. Oh, this is great. We'll be done in six weeks. No. No. You better have a two hundred thousand contingency fund here to pay for your penalties. Right? This is serious stuff, and and I wanna say the underlying technology is quite old. Like, the light bulb was invented long, quite a few years, really many years before there was a power grid, right? One of our key problems here is that people are scared of this stuff. And I know that within HCSS, they talk about fear, uncertainty, and doubt, FUD. I tend to use the scientific term for that, which is post traumatic statistics disorder or PTSD. So we have to be careful to avoid any of that and just talk about simple things, like chances. Now let's do a cost estimating model here. And so you're estimating eighty thousand for sheet rock, hundred fifty thousand for electrical, blah blah blah. It adds up to six hundred fifty thousand dollars are your estimated costs. And say, well, okay. We'll bid seven hundred thousand. Well, the cost can be higher than six fifty. Do do you have a freaking clue what your chance is of losing money if you make this bid? Would you like one? Because we can help out. And now we're gonna go to where the data comes from. And, again, this is where Aaron was entering estimates on his own. If you keep, you know, kept track in your work in progress report, your whip, we're gonna go from whips to sips. What are the sips? The die rolls, but not a real dice. Okay. We're looking at paving, job one, job two. Here was an estimate. Here was an actual. Here's the percentage difference. Right? And and one of the one of the measures that Aaron was was using, he he he gave three ways to measure risk. They're all completely valid. One of them was like the percentage overrun in effect. So if I go back here now, we have in effect created the die rolls of cost overrun. So here's your percentage, the percentage you are off. Right? We're not in Kansas anymore. Could be that. Could be that. Could be that. Could be that. And the chance of exceeding seven hundred thousand is forty five percent. Well, wait a minute. What if we bid seven twenty five? That is thirty two percent. What if we bid a million? It's zero, but you'll never win the bid. So what should that number be? That's above my pay grade, guys. But at least you should be making a chance informed, decision on this stuff. So where do we go from here? Again, I was I had no idea. I I have a friend who now works at HCSS and got to meet these people and had no idea they had this infrastructure in place, which I think we could view as a giant power plant and then sort of create a power grid of of probability, beyond that. So I kinda mock that up. Now here we have an example where, presuming that you've got enough data in the system, enough past experience. Right? So here we bid a million in labor, million and a half in materials, five hundred thousand in extras. That total is three thousand three million. If you bid three two fifty, you have an eight a twelve point six five percent chance of loss. And here, let me just take you out to the, the real deal here, which I think I have up running. Let me zoom in on this. So this is an actual this is not hooked to real h CSS. Okay? But I run down here and change my bid to three million three hundred, and it goes to eight point five five percent. So I think all the pieces are in place here, and and the nonprofit has developed really exciting new open standards and technologies for for moving chance and form data around. So I'm I'm really looking forward to, you know, to customer interest in this and the ability to start really hooking up the heavy bid power plant, to dashboards like the thing I just showed you. Alright. So that basically leads us into where do we go from here? So you've heard from doctor Savage. You've heard about chance count. You've heard about the the flaw of averages, and you've seen in heavy bid what we can currently do. So the question is, where do we go from here? I'm gonna steal back the screen here. And we have basically been talking about, you know, manually entering data here, production rate ranges, best case, worst case scenario, the cost ranges. The the problem is that's a lot of manual entry on top of entry that you're already doing in your estimates. How do we automate that so that you can take the information that you already have and make better use of it. So some of the things that we've been thinking about, utilizing something that the majority of you are already utilizing. What about something like production history? As doctor Savage, showed in his demonstration with the work in progress image, you have the jobs that you've done in the past. We know exactly what our production rates are on those jobs that have that particular item on it. We can use the production rates from those jobs to help us better estimate our chance of risk for any given item. The other items that we can look at are, you know, bid results. Bid results in the green sheeting. I know Paul is gonna be working on a project here, very shortly to take the the green sheeting report that we have in heavy bid and take it to the next level. We can use all of that information to track not only our historical results, but also our actual productions out in the field to not only make this more accurate, but make it much more user friendly. We also have the ability to bring in outside data. We can bring in, safety data. We can analyze our safety risk. How many incidents do we have on any particular item? Do we need to account for additional dollars because of potential safety hazards that we have in, on this particular type of work? So the question that doctor Savage and I have is, is that something that we should pursue? And if it is something that we should pursue, where do y'all want us to pursue from? Do y'all want us to look more at bid results? Do you want us to look at, production data? Do you want us to look at something completely outside of it such as safety data or something else? So at this point, I'd like to open the floor up to to some q and a and get some feedback from everybody. So we've got two questions related to production data. So, yes, it does seem like production data would be the place to start. We already have it, and we can gather that data and bring it in. We can utilize that to help populate some of the the risk items in there. And where would that data come from? So the screen that I have up for you now is what's called production history analysis. That data pulls directly from your heavy job data. So as soon as a time card gets submitted from the field, it updates, this data. So it doesn't require any kind of manual intervention from the end user. It's all done in the background. And I wanna mention something about this. Let's remember the boss also has PTSD, and you don't wanna go sticking your neck out on predicting uncertain things. But if we've got it straight from the heavy bid database, right, You're saying, hey, boss. The last fifty times we flipped a coin, it came up half heads and half tails. Or the last fifty times we bid on a paving job, we were ten percent over. Right? You don't have to state this as your opinion. You just say, if things don't change, we have a thirty percent chance of being over budget by a hundred thousand dollars. Right? So no one has to stick their neck out here. Absolutely. And, Chris and Kevin, to answer your questions, we absolutely do have a place to enter in historical data into heavy bid that is not tracked through heavy job. Production history is tracked through heavy job, but we do have a function, called cost history that allows you to manually import historical actuals. And you can import that via an Excel import and track your productions, from whatever system that you use to track. Obviously, production history is an automatic process. This will require, manual upload, but you can still get, very, very similar results. Now bear in mind also that because there's a power plant here that can generate the electricity, we could take the SIPs out of heavy bid, and you could run them in Python or R or just in a spreadsheet, which means the the boss does not have to be within this system. You don't have to take the boss down to the power plant. You could just email the boss's spreadsheet and say, hey, boss. You change this cell. So it's a million bucks right now. Says we have a thirty percent chance of exceeding it. You change it to a million too. See what it's gonna see what the chances are. The boss can do that on on their own based on the heavy bid data. You're you're not gonna make up the data on your own without heavy bid, unless you got statisticians working for you. But, again, it's like a division of labor and may reduce the the boss's post traumatic statistics disorder. If the boss doesn't have to be sitting there with a bunch of cohorts, well, you understand this boss, you understand this. Bosses don't like that. We've got, kinda looks like we got two questions from you. First question is, what is the largest risk in using Monte Carlo versus SIP? Monte Carlo versus what? Versus SIP. Oh. Oh, wait a minute. Wait a minute. No. No. SIPs are Monte Carlo. They're just stored Monte Carlo. Okay? I'll give you another analogy. If probability were beer, the SIP is a beer bottle. It is a way of storing the Monte Carlo results in an auditable way. So it's just a way of transporting Monte Carlo. Again, I go back to, do you wanna go down to the power plant at night to read your book, or do you wanna read it at home? Just a way of transporting data from things like the heavy bid risk, analysis that that that Aaron just demonstrated. And and by the way, I mean, I think the risk analysis, all the infrastructures in there, It can be tweaked and refined and calibrated and all those things, but it's still the basic same infrastructure. And and, then once you can get the SIP libraries out to the users, they they can they can query this in a much friendlier private environment. And then, I see the big difference is that we were able to use actual production data within the Monte Carlo simulation. So that's definitely something we are looking at accomplishing in the future. We wanted to make sure that that is where we where our customers, y'all, want us to start, start expanding. Kevin, to look at your other question, the standardized activity database that all users can opt in with is definitely something that we have explored in the past, but we have not made a lot of headway on that just simply because from what we have found, everybody likes to do things their own way. So trying to make it work has not been something that we have found really a clear path forward with. It's definitely something that was the door has not been shut on, but it is something that it's the best way to say that it's not something we have a whole lot of traction behind at the moment. Aaron, I wanna comment. I I just wanna make another comment on the historical data. We haven't really discussed it in detail, but one of the things that the nonprofit has done aside from figuring out how to distribute, you know, the probability, is way much better ways to fit that data, making it much easier for you internally to use that historical database. It's technical. I don't want it into the weeds now, but I hope in future discussions, it will make the whole process of of using your historical data a lot more seamless for you and for the customer and for everyone involved. Kind of a technical mathematical detail that we don't know we don't need to go into the don't go into the weeds on right now. Fully fair. Looks like we have two two hands raised as well. Tristan, Sarah, I'm happy to allow you all to talk. If you want to, you can either talk or we can communicate through, through the chat, whatever you prefer. Tristan, I'm gonna go ahead. I'm gonna let you talk. That way if you wanna communicate, you can unmute yourself and communicate that way or, communicate through through the chat or the questions. Sarah, no worries at all. Alright. So we got another question here from Christian. Can risk profiles be saved within the master or the codebooks? A master profile that would hold mass attributes, for typical items or the activity codebook having the attributes held within the codebook. So having the having the risk, risk assessment, so productivity risk, cost, or cost factor actually save within either activities within the master estimate, which we can already do because it's nothing more than a template estimate, but storing them within the activity code book. So if we were to go to the activity code book, basically adding another category somewhere below the calendar production rate in the report groups where you could enter in the risk information. The does anyone else feel like that would be something that they would find very they would find useful and would wanna implement in their own environment? You can go ahead and raise your hand or, simply reply yes or no here in the chat. Yeah. Into the Centimeters master's program, Arizona State, Oregon, see if a grad student could assist in the thought process development with their thesis. I hadn't considered that. That is that is a very novel idea. I'll dig into that one. Arthur chiming in. And, Arthur, I tend to agree with you as well. Kyle. I don't I don't mean mean to laugh at your question. I saw, Christian's response. He just doesn't want us getting bored. I appreciate that. Kyle, to answer your question, absolutely. This recording will be available, in the next few days. I'll, it's gonna be on help dot h c s s dot com. If you search monthly webinar May twenty twenty four, this will show up, typically in about a week, and then your colleague can view at their convenience. Kevin, I agree. That is a that is a great idea, and it gets an outside perspective as well that, you know, somebody that might not necessarily have tunnel vision can can give their their thoughts and ideas around as well. Aaron, wanna repeat that one for me? Kevin Kevin's thought again. Kevin, talking about having a master's a master's student get involved in the thought around, standardization with with the activity database and getting getting their input on how all of that would how all of that could, potentially work. Yeah. So because one of my master's students just did that, and we should put our heads together and get our master's students together on this, as a matter of fact. But my guy is actually a captain in the army, who, you know, who's who's gonna be involved in construction. He's not in the not in the army corps of engineers, but he's a real smart guy in, you know, civil and environmental engineering at Stanford. And I would I would love to, to collaborate with others, you know, at other universities and things on that. That that'd be a very interesting thing to do. So you can put us in touch, Erin, or or or do whatever. But, I mean, heavy beta seems like a beautiful place to start this analysis. Absolutely. Kevin, yeah, we're gonna I'm definitely gonna dig into that, and I appreciate the the heads up on that. Everybody, thank you for your time. I know we're running a little bit long. I've got two, two questions, one hand raised and one question. So we're gonna go ahead and answer that, before we conclude the meeting today. Doctor Savage, I'm gonna let you answer this one. Chris asked you what your favorite dessert is. What my favorite dessert is? Yes. Pecan pie. Why? I know. You might have something show up in the mail. I don't know. And then, anonymous, doctor Savage, thank you. This was a great, you're a great teacher, and I've enjoyed this webinar. So, kudos to you, doctor Savage. Wow. I thank everyone for coming. And then, Chris, I'm gonna go ahead. I'm gonna allow you to allow you to talk. So if you would like to unmute yourself or if you'd like to reply, you're welcome to ask whatever questions, you would like at the time. Alright. Looks like, Chris, put his hand out. Everybody, I wanna thank you for your time, doctor Savage. I think this webinar was fantastic. Thank you for for donating your time to to come out and chat it up with all of us, and sharing your knowledge of what is a a very, very involved topic. So, everybody, thank you for your time. Everybody have a great day. And if I don't hear from you today or tomorrow, have a great weekend, and I will see you all at the next one. So long, guys.