[00:00:00] Gary Fischer, PE: Next we welcome Ivan Damnjanovic who is a professor and director of engineering, project management at Texas a and m. He’s also another great collaborator with PPI. Ivan specializes in qualitative and quantitative methods for management of engineering and pro project risk, as well as management of infrastructure and transportation systems.
[00:00:25] Gary Fischer, PE: Has extensive engineering experience and engineering risk and safety analysis applied to projects from all kinds of different in industry segments. Yvonne, welcome and let’s hear about what you’re doing with production modeling based risk management.
[00:00:41] Ivan Damnjanovic, PhD: Thank you Gary. Can you all hear me well?
[00:00:43] Gary Fischer, PE: Yep. Can hear you and see your screen. Excellent.
[00:00:46] Ivan Damnjanovic, PhD: Today I will be presenting about the research and educational initiative at Texas a and m University that deals with project production modeling. The research effort that we have which fits into this larger effort of project production modeling is in variability and risk.
[00:01:09] Ivan Damnjanovic, PhD: Here we take the more of a aleatory viewpoint on this variability being that is independent of the observer. As opposed to what epistemic uncertainty would be, that would be more of a knowledge related uncertainty. So things are, in our world, random in spite of our efforts to control all the conditions.
[00:01:30] Ivan Damnjanovic, PhD: So if we run the experiments with all conditions being the same, will be obtaining the random random observations out of this variability and randomness risk events have emerge. So risk events are typically materialized. As a result of this variability now as a true scientist, I’m gonna start this presentation with an excerpt from the Old Testament the book of Ecclesiastes by King Solomon, which kind of highlights the same sort of viewpoint on variability.
[00:02:06] Ivan Damnjanovic, PhD: I’ll read it for you in return, and I saw under the sun that the race is not to the swift. Nor the battle to the strong, neither yet bred to the wise nor yet reaches to men of understanding, not yet favor to men of skill, but time and chance happen to them all. In a way, our research focuses on time and chance that happen to the projects.
[00:02:33] Ivan Damnjanovic, PhD: So let’s look at into the two areas we are currently working on. So we are currently working on contingency time, cost, or materials. The things that are buffering, variability. And the second thing is we are dealing with changes. These changes typically happen throughout the project. Many people will say, we’ll deal with the chances with a hundred percent probability the changes will happen, and we’ll need to understand how do we deal with that?
[00:03:01] Ivan Damnjanovic, PhD: So let me go and start first with the, first topic, estimating and managing contingencies. So we use schedules for setting up contingency at the various level, but we all know how this works. Basically, we have a critical path network. We find a critical path, and from there we can either do this analytically or we run some form of a Monte Carlo simulation.
[00:03:27] Ivan Damnjanovic, PhD: Some advanced methods, let’s say it would use, Dependency between the paths, dependency between the activities. So they will be competing paths and ultimately we will get some distribution of times and we’ll try to set up the contingency within certain level of confidence. And this is how typically it works.
[00:03:49] Ivan Damnjanovic, PhD: However, this. We all know falls apart when we introduce resources and the constraint of resources. So if you have ever worked on a resource load schedule, you know this all too well. So the floats automatically disappear as the re as the resources are being shifted from one activity to another and the path, the critical path rearranges.
[00:04:14] Ivan Damnjanovic, PhD: So that is the big fault. The big, problem if we ever start to estimate contingency here. Now, variability also. It’s being underestimated here because variability starting to affect, in some cases the throughput or the cycle time we need to be doing we need to have in our, project.
[00:04:36] Ivan Damnjanovic, PhD: So this is all missed all together in a critical path method and ultimately ignores the work workflow, which really means we don’t see what flows through the network so that we can easily control it. Now in past. Contingencies and risk estimation was again, based on the method of moments. This is analytical method that relies on finding finding higher other moments of functions of random variables.
[00:05:07] Ivan Damnjanovic, PhD: And Monte Carlo simulation, which does that through a simulation. And we are shifting more to analytical methods, and those methods are the methods where we. Assume certain things. We assume type of q we also can advance from there and go into more analytical metrics, methods where we have less assumptions.
[00:05:30] Ivan Damnjanovic, PhD: And ultimately we can use discrete event simulation. Our mind shift in estimating contingency and managing contingency, which is also an important aspect, shifts from more schedule based into production based. So that is one aspect of, the research we are working on. Now it begs another question here.
[00:05:54] Ivan Damnjanovic, PhD: Do we really like variability? Okay, I know what you’re gonna say. It does increase contingency cycle times working process requirements also decreases to some degree throughput and as well generates condition for risk events to materialize. So in short, we’ll say really no, but let’s look at few, things here.
[00:06:24] Ivan Damnjanovic, PhD: Let’s compare these two technologies. Assume that you are in the process of of adopting one of the technologies concept design, and you have technology one with, a mean value, expected value of about $9 million. And then you have technology two that that is more expensive, about a thousand dollars, a thousand.
[00:06:45] Ivan Damnjanovic, PhD: A million dollar. Okay. More expensive. And assume that you as a project manager will be will be judged based on the performance of your project. And you can either be a hero in the, in case the project ends up to being less than 8 million or a goat. And I, by gold, not the goat of greatest of all, but in all fashion term goat as scapegoat and if the project ends up to be higher than $10 million. So technology one probability of goat is 16. Probability of goat in technology two is 50%. So higher probability of being goat lower mean for technology one, the hero there is a slightly higher pro chance of being hero. Actually double, almost double because that technology too is associated with higher variability.
[00:07:42] Ivan Damnjanovic, PhD: So regardless to variability, we’ll say, okay, technology one makes sense, but how about if this choice is optional and can be made later? So if we do a little math here I’m not gonna bother you through all of this, but probability of goat if we consider both technology at the same time, can be reduced by, half, basically 8%.
[00:08:08] Ivan Damnjanovic, PhD: And probability of being a hero can be moved. To 42%. So carrying out design alternative in, in projects provides a value. So the more valuable these alternatives are, the higher the value it provides. But the key here is that we like the variability. When can, we can take only the good parts. Not the bad parts.
[00:08:38] Ivan Damnjanovic, PhD: If we take this optionality or ability to pick and choose from our from our situation here, we may have situations like this. So if we are not, we have no ability to pick up only the good parts. If we are generally in the top, we love it. We are at the good position. Okay? So we are in the position where we can only go down.
[00:09:05] Ivan Damnjanovic, PhD: Then our variability is something that we wanna avoid in a complete opposite, if you are at the bottom, okay, we are embracing. Variability because variability at that point can only take you to a state which has a higher value to you. Okay? And this is very natural. And you can see people say desperate people make desperate choices.
[00:09:35] Ivan Damnjanovic, PhD: And, this is what, variability does it variability. Prompts. We seek out variability when we are in desperate situation. However, most of our life and most of our project, we are not really in those two extremes. We are not in the top and we are not in the bottom. We are somewhere in, in the middle.
[00:09:55] Ivan Damnjanovic, PhD: So we often see the danger and the opportunity in variability, but only if we consider that variability as optional. And that brings up, brings me up to the second type of research that we are doing. When we see the variability and the value it pro it can provide. So I’m talking here about the design changes and responses to new discoveries.
[00:10:21] Ivan Damnjanovic, PhD: Okay? This kind of goes into the very similar to what lean, agile principle is. Assume the variability, preserve the option, and respond by last responsible moment. We can value this using a real options theory. And there has been a lot of methodology introduced with various conditions and, constraints of how would you do that?
[00:10:48] Ivan Damnjanovic, PhD: So let me now go over a very simple example and I’ll assume that there is some design. Design sort of process that is integrated with construction and construction starts when the design is not fully complete and let’s say consider three different alternatives. Design options of some element that we are gonna have to deal with.
[00:11:11] Ivan Damnjanovic, PhD: Alternative A, which could be a bulletproof solution where no associated risk with it. Let’s assume that it’s, basic, a baseline. Skin doesn’t exist, but we have it here anyways, just for illustration purposes. Okay? Assume there is. Now alternative B is a good solution, but there is a risk associated with this.
[00:11:33] Ivan Damnjanovic, PhD: Alternative B if risk event B occurs, the price of the project or duration of the project, or whatever performance metrics you wanna consider, the pro the alternative B increases and is higher than alternative a. If it doesn’t, then either stays the same or it goes lower in this case.
[00:11:56] Ivan Damnjanovic, PhD: Okay, so we have the expected value, alternative B to be lower than the alternative A, which is a good solution, and then we have a risky solution. Okay? On top of that, we have three alternative, risky solution. Risky solution possibly the best, possibly the worst. Okay. If risk of an zero occurs. The price of alternative C is alternative C plus, which is all the way there otherwise is actually lower than that.
[00:12:25] Ivan Damnjanovic, PhD: In our conventional logic, we will go with alternative B. We’ll select this alternative. Sometimes the, maybe even lock in this alternative freeze it, freeze as a design and add some contingency to take into account possibility of something going wrong with that alternative. By that meaning that it, we should move potential, we could move potentially to alternative b.
[00:12:52] Ivan Damnjanovic, PhD: Now if we choose to still consider three possible design alternative during this discovery process, okay, alternatives, B and A, so which will be now purple. So if we combine two alternatives, we have something like this. Okay? If risk occurs. We don’t go all the way to B minus, but we are limited by the value or the cost of alternative A.
[00:13:18] Ivan Damnjanovic, PhD: So we are basically basically creating as asymmetry in this variability. So we are taking only the good parts of alternative B and limiting bad parts of the alternative a, alternative B with the alternative a. Or if we consider all three of them, which would be now this green line, we can even get better expected value.
[00:13:45] Ivan Damnjanovic, PhD: Okay? Now, in the reality we have this phenomenon where these co this values that you’re getting reduction maybe not necessarily reduction or cause, but prevention in further escalations of the cost. Okay? But for the sake of simplicity, you can see now that alternative A, alternative B, is greater than alternative B and A, and it’s a greater than alternative B, a, and C.
[00:14:15] Ivan Damnjanovic, PhD: And we wanna select the smaller one. Bringing up design alternatives throughout the construction process actually improves through discovery process actually improves value, but it comes with a cost. Okay? There is a cost and where is this cost coming from? It is the WIP. Okay? How much such change will cost you?
[00:14:39] Ivan Damnjanovic, PhD: Depends on a WIP. Just walk around the construction site if you wanna measure the risk associated with design changes. Now, walk the construction site, see what you know, casted concrete, precast precast, prefabricated element purchase equipment, but not only on the construction side. Look at also the supply chain.
[00:14:58] Ivan Damnjanovic, PhD: Look at what you have as engineered toward long delivery times. Anything that is currently flowing through them. It is your WIP, basically. Your WIP is your exposure. Okay. It’s exposure to potential to potential design changer. We may not know the probability of that event happening, and of course, probability of such event.
[00:15:25] Ivan Damnjanovic, PhD: Again, many people will say it’s a hundred percent that something’s gonna happen. How bad will see, okay, we don’t know the probability, but we can manage the risk by managing exposure to this. Now, not all of the WIP. Is actually is going to be your exposure, but we, know that some of it will be so WIP is your total exposure and the actual consequences is gonna depend on portion of that WIP.
[00:15:57] Ivan Damnjanovic, PhD: That needs to be changed still. We need the right amount of WIP to get the work done to ensure the right amount of throughput we need. We need to minimize the cycle time and exposure to risk. It’s the design risk, the risk of design changes, but we also need to maximize the throughput. It’s really a balancing act and some.
[00:16:17] Ivan Damnjanovic, PhD: Folks like scientists will say that’s the optimization. You need to understand the production system in order to optimize it. So we are currently working on the maturity models that project teams could use to assess if the level of understanding of the proposed project production system, it’s sufficient enough to make these kind of decisions and manage the risks.
[00:16:45] Ivan Damnjanovic, PhD: Here is the WIP, of carrying alternative B into alternative A, the WIP of moving through the risk event number C. So all of those are basically your exposure. So managing WIP, it means managing design change risk. So that is in short the, two research ideas we are currently working on that relates to project production modeling.
[00:17:17] Ivan Damnjanovic, PhD: Now I’ll be talking a little bit about our educational effort and that’s the effort. Texas a and m does jointly with the project production institute. Here. The program is reinvent the way we build, and it has been now for almost three years, I believe we’ve been working on this. All the certification pro process start with a basic two day course where we introduce the key concepts of pro project production, modeling, production systems variability, five levers.
[00:17:52] Ivan Damnjanovic, PhD: We talked about all of that is critical. For you to move to the next stage of actually implementing next gen project management methods. So it’s a two days. We teach it typically in Houston. It is right by the medical center, so it’s it’s quick access to, Houston downtown. So it’s a two day course.
[00:18:14] Ivan Damnjanovic, PhD: This is some of the photos that we have from the previous meetings. We just finished one in November, and I’ll show you some. Future dates in just a second. What is this certification program? As I said, it starts with this introductory course. So it’s a two day in-person course.
[00:18:33] Ivan Damnjanovic, PhD: Following those course, you qualify to take the next stage and which is, the three day in-person. Certification. Okay. What is that? Three day three days in per in-person courses. Those courses focus on either engineering, construction or procurement. We often group them. What we offer is engineering construction together.
[00:18:57] Ivan Damnjanovic, PhD: And that is, that is quite useful because many of our many of the people who are looking for the certification work in both areas. So after you complete that three days, so five day altogether, you qualify to actually work on the six to 12 months capstone. Projects. You can take one Capstone project or you can take two Capstone project.
[00:19:22] Ivan Damnjanovic, PhD: We have yearly cohorts. We have finished one cohort and Keith is gonna be talking about that in just a little bit. And ’cause he’s a member of the previous year cohort, and we’ll start in January, the new cohort of students that will be working on students by same students, meaning professionals working on the, Optimizing their production system. So if you do that capstone, you automatically qualify for a specialty certificate in construction. If you do construction spec, specialty certificate engineering, if you do engineering and by completing both, you automatically qualified to get the master certificate in PPM.
[00:20:07] Ivan Damnjanovic, PhD: So join us in our next project management educational initiative. You will learn the fundamentals of next gen project management production system. You’ll learn by doing, by bringing up Capstone, by bringing up real project through Capstone project and working with us. You’re gonna be studying with the student cohort, the.
[00:20:31] Ivan Damnjanovic, PhD: Professionals like you as well as we, will work with us PPI and Texas a and m instructors. You will ultimately earn professional certification, whether it’s a specialty or master’s. And finally, you’ll you’ll establish a community of practice and you will work in this now. Educational institutions like Texas a and m, we are trying to bring more and pro more of project production modeling philosophy and, examples and models into our into our classes.
[00:21:07] Ivan Damnjanovic, PhD: So both in our undergraduate classes, we start talking about the production systems, not at the same level as we are doing later in our graduate classes, but there is a, concentrated effort to bring up. Project management production focus. In curriculum, in in our general civil Engineering, engineering pro and engineering project management courses.
[00:21:34] Ivan Damnjanovic, PhD: So here is the last slide. It’s about the course offering there in Houston. So we’ll have one scheduled in February, and that one is gonna be focused on data center. So if you have, a project that may deal with data center. If you’re excited about data centers, please join us for that because that one is gonna be about data centers and it’s gonna be still qualifying you to go through the certification, but it’s gonna have that more of a flavor on data centers, one in May, one in November.
[00:22:04] Ivan Damnjanovic, PhD: Those are all two days. And combined courses we have is next gen, engineering and construction, one in May and the one in November. With that, I’d like to thank you all for the attention and I’m open for any questions. I’ll bring up this Aggie thing here, standing together for a bigger impact.
[00:22:24] Gary Fischer, PE: Alright, thank you Yvonne. Okay. Questions from our audience for, Yvonne? I’ll get the ball rolling. When are we gonna have that maturity? Model tool?
[00:22:36] Ivan Damnjanovic, PhD: I’m hoping sometimes by the end of the spring, early summer, we are gonna have some ways of judging how much we actually know, do we have the right sort of information so we can so we can start thinking about what production what production system we can develop and how to model
[00:22:58] Gary Fischer, PE: good.
[00:23:00] Gary Fischer, PE: I’m giving the system a little bit minute to refresh here ’cause it’s some latency in the questions.
[00:23:06] Ivan Damnjanovic, PhD: Oh yeah.
[00:23:12] Gary Fischer, PE: So I guess your risk thesis there is hypothesis is really pushing decision-making out as far as you can responsibly. So that you keep your options for a different alternative or approach or whatever it is a choice. Open as long as you can. So don’t make decisions before you need to. Is that the bottom line of what Correct.
[00:23:41] Gary Fischer, PE: Get across?
[00:23:42] Ivan Damnjanovic, PhD: Yeah. No, so that, that’s one point is don’t make your decision until you need to carry on. You design alternative options. You manage your risk by carrying out design options. Okay, now that brings up that’s, I wouldn’t say that’s something dramatically new, but even in the Agile philosophy, we are talking about last responsible moment real option theory.
[00:24:08] Ivan Damnjanovic, PhD: We will talk about the value of waiting, wait and see rather than just commit right now. But what is different here is because we often didn’t think about this from the perspective of risk. Managing the risk associated with design changes and WIP the exposure because we get, we knew about the value and I think for a long time this has been well established, but people don’t always think that that there, there will be some consequence.
[00:24:38] Ivan Damnjanovic, PhD: They think about, there is a consequence of design changes, but they don’t know what the consequences are and the consequences are the WIP. So if we think about the risk as probability times the consequence. We know what the consequences, we can manage the consequence, even though we may not know what the probability is.
[00:24:59] Ivan Damnjanovic, PhD: So I think that’s the second layer to that to that value of. Bringing up design alternative.
[00:25:09] Gary Fischer, PE: Yeah. So are you creating a white paper with this?
[00:25:12] Ivan Damnjanovic, PhD: Yeah, so we are, actively working on this great. We have a student that is working with me and and we have done something in past, already along these options.
[00:25:23] Ivan Damnjanovic, PhD: Now we are trying to illustrate this with the analytically traceable model, so we can say that, hey. This is it. Okay. You we are trying to prove something,
[00:25:35] Gary Fischer, PE: but I think our audience will be interested in reading it when it’s published. We’ve got a question around that. Yeah. We’ve also got another question.
[00:25:40] Gary Fischer, PE: How are you estimating the WIP cost for the alternatives?
[00:25:45] Ivan Damnjanovic, PhD: Yes. So basically, so when you are bringing up the design alternatives, you are also carrying some elements of design process. So basically your cost of carrying all design alternative needs to be. Compared to the value of carrying those alternative.
[00:26:04] Ivan Damnjanovic, PhD: So there is a cost. That is a great question. So there is a cost associated with carrying out the multiple design alternatives. So it is now in, in reality, what happens is we, see the need for design change and we. Even if we didn’t think about the design we, think about the fixes. So this is a fix.
[00:26:28] Ivan Damnjanovic, PhD: This is a fix. What, we are gonna do if something goes wrong. Now the, that fix could be cheaper or more costly. The cheaper version is if we properly manage WIP, and the more costly version is if we. Have loaded our system with with inventory and our cycle times are longer and we have WIP that we need to throw away because of that.
[00:27:01] Ivan Damnjanovic, PhD: Timing it it is it is, a timing decision of when do you wanna, how long you wanna carry those alternatives from the perspective of the cost and how much, value you gonna bring.
[00:27:17] Gary Fischer, PE: Great. We definitely interested in your paper or is it gonna be a thesis? Is it
[00:27:23] Ivan Damnjanovic, PhD: yeah, so it’s gonna be a white paper first, but ultimately it’s end ends up in a, as a thesis.
[00:27:28] Ivan Damnjanovic, PhD: Yeah.
[00:27:28] Gary Fischer, PE: Excellent. We got a lot of interest in reading it when it’s published here, so we’ll make sure we get it out to our audience when it’s published.
[00:27:35] Ivan Damnjanovic, PhD: Excellent. Thank you. All right. Very
[00:27:36] Gary Fischer, PE: good. Thank you,
[00:27:37] Ivan Damnjanovic, PhD: Yvonne. Thanks.
