What’s Going to Happen on My Project?

[00:00:00] Roberto J. Arbulu: How do we make sure that our employees are actually on time? All right. And so we will explore together with Chet, I’m going to take you through an introductory part on some key ideas. And then Chet will take us through some examples of what other companies are doing to be able to answer this question on what is really going to happen on our projects.

[00:00:21] Roberto J. Arbulu: Okay, I want to start with something I found the other day on the web and really took my attention. I’ll let you read it.

[00:00:34] Roberto J. Arbulu: Okay, so I don’t know who KL is. All the credits go for. To KL, but that’s what we see. Okay, our professionals. We are spending a lot of time on this left queue. Trying to, you know, explain pretty delays, right? And the question is, can we move to the other queue, right? What are we doing to avoid them, right?

[00:01:04] Roberto J. Arbulu: What techniques, what thinking is behind? All these. Okay? And, so I just wanted to start with something hopefully a bit funny. Okay? Because the next news I bring to you are not that funny. Okay? Sorry. In 2013, our friend from McKinsey, and where is Jim Banaszak? Hey, Jim. Publish this report, okay? I’m not going to take you through all the details, but basically, what you’re seeing is that 90 percent of the projects incurred cost overruns and delays, the average cost increase is 80%, and the average sleep is actually 20 months behind the original schedule, and this looks at mining, oil, and gas infrastructure.

[00:01:54] Roberto J. Arbulu: If you are the lucky 2%, please raise your hand. Okay? But the important message here is what they have done. Last year, okay, almost 10 years after, okay, and the situation across multiple industries, you know, waste and water, railways, is not really changing much, okay. They analyze about 500 or so projects, roughly, and we’re talking about significant delays, significant overruns, okay.

[00:02:32] Roberto J. Arbulu: And so, There’s something happening. Things are not changing, at least, fast enough. Right? At least fast enough. Okay? And so, The question is, with this data, do we know the answer to this question? Can we get the answer to that question? Right? Do we, do we want to be another dot in those statistics? Right? So, What is also happening, and I don’t know if you’re familiar with this idea of the ability to influence, is a sort of 1975 old schematic that tells us that the influence basically goes down as we move the project, right?

[00:03:17] Roberto J. Arbulu: Very well known. What we see, because of that, is that management intervenes at the end. And I think Gary’s, your story are basically an example. With obviously much more detail of what this means, okay, one of the things that you will see as we go through is that one of the reasons, not the only one that this is happening is that as we make decisions throughout the project life cycle.

[00:03:48] Roberto J. Arbulu: The decisions that we make, right, are not helping us to raise variability, and also they increase work in process, which is what we mean by WIP, quite significantly. And we’re going to explore and explain what this really means. So keep, keep that thought as we go through, okay? But if this is all happening There should be some gaps.

[00:04:12] Roberto J. Arbulu: We’re not here to do a forensic analysis of what all the different gaps might be. I think Gary already alluded to many of them. But we want to bring your attention to one particular common denominator that we have found as we have worked across multiple projects around the world. Different type of projects.

[00:04:31] Roberto J. Arbulu: And is this idea Okay, not an idea, excuse me, but a fact that in our industry, in engineering and construction, in the project world, we put a lot of emphasis on project administration. Right? We do contracts, we check progress, right? We do a lot of that. And we have an over reliance on reporting and forecasting of project progress, right?

[00:04:57] Roberto J. Arbulu: We do a lot of that for this, okay? But unfortunately, what we don’t necessarily focus enough, that’s what we’re saying here is an underinvestment, is on how we produce. I know this sounds simple, right? But if we want to improve the performance of projects, we need to focus, we need to touch the work. We need to focus on how we work, how we produce.

[00:05:21] Roberto J. Arbulu: On top of many other things, of course. Okay? So today, we’re going to explore What we mean by these and what are the key ideas behind, you know, increasing our investment of efforts, time, attention to the way we produce. So, production management, this is one of the key ideas of our presentation, is that it provides the means to ensure predictable outcomes.

[00:05:47] Roberto J. Arbulu: We will share some examples of people like all of us, right, professionals, industry, companies in the industry that have been able to achieve this. Over time. These examples are just some of them, right? Companies that are actually able to compress schedule by a significant amount of time. You know, 40 percent faster, 47 percent less cost in oil field development.

[00:06:18] Roberto J. Arbulu: LNG facilities in oil and gas that have been able to reduce cycle times by 13, 52%. becoming a benchmark, becoming much more predictable, significantly much more predictable, as well as a significant reduction on cost. Okay? And these are just four examples of things, of projects that are achieving quite significant results.

[00:06:43] Roberto J. Arbulu: Okay? The question is, what is the thinking behind what they do? Okay? I’m not going to talk about what exactly they do, but what is the thinking behind what they do, right? How are they looking at the challenge? The first and most important idea, okay, and we can probably spend one day on this slide alone, is that schedules are not representations of production systems.

[00:07:13] Roberto J. Arbulu: This is a very, very, very important idea, okay? It might not be a new idea to you, but we need to talk about it, right? Our belief that the schedules represent production systems is flawed. Okay, and we’re going

[00:07:28] James Choo: to expand on that.

[00:07:31] Roberto J. Arbulu: So, from a production perspective, Once again, Gary mentioned this, the schedules represent, They’re a very important tool, if you will, Because they tell us what is the demand that the production system should respond to.

[00:07:47] Roberto J. Arbulu: Okay, so let me give you a simple example. If I have 10, 000 piles, right, that I need to drive. And I have 10 months to do it. That means I need to do 1, 000 per month. Let’s say 20 days every month. I need to do about 50 every day. Right? My production system should be able to generate 50, should be able to produce 50 every day.

[00:08:19] Roberto J. Arbulu: Okay, and so when we say demand, demand is a rate. Okay, and we use schedules as a means to establish these rates. Okay, what is the production system, what should the production system respond to? If the production system doesn’t have the capacity to produce to the scheduled demand, we will experience scheduled delays.

[00:08:42] Roberto J. Arbulu: Okay, there’s a whole production system on the right. Okay, that has certain capacity. And if that capacity, I repeat, doesn’t meet the demand, we will experience, we will experience scheduled delays. If the production system has the capacity to meet or exceed that demand, we might have an opportunity to accelerate, and depending on the resources, we might also have an opportunity to reduce cost.

[00:09:10] Roberto J. Arbulu: Okay. And so what these organizations are focusing on is understanding the right side of this equation. Okay, the right side of this equation. More scheduling doesn’t necessarily give us better predictability. More detail in our schedules doesn’t necessarily give us more predictability, right? And so one of the key thinkings, as I said, as I said, is we’re going to spend a lot of time on the right.

[00:09:42] Roberto J. Arbulu: This is an schematic actually an analysis, a real analysis on a project that what you see on the left is the demand as projected, if you will, by the schedule, right? But when we started analyzing the production system, we learned that the production system doesn’t have the capacity to produce that demand.

[00:10:04] Roberto J. Arbulu: And it’s gonna be. Sort of limited, right? Therefore, we are going to have an issue with our schedules, right? And right there, without doing much more analysis, we know that there’s a high chance actually we’re not going to make it, right? And we will become one of those dolls in the statistics, right? And so one of the first steps that organizations are taking is do the analysis of the production system and see, I repeat, if the production system is able to meet the demand or not.

[00:10:33] Roberto J. Arbulu: Okay? So we’re not saying get rid of the schedules or anything like that. We never said that. We are saying, let’s use them, but to understand that demand. Okay? And so the schedules and the production system are really two gears, right? If we want to move faster on the schedule and the production system cannot, it’s not going to let us.

[00:10:58] Roberto J. Arbulu: Right? The two gears will not move. Right? If the production system can do more, We can move it, and maybe we can accelerate the other side, and they will actually achieve some significant result. So keep that concept in mind as we go through this. So, from a production perspective, the idea of a production system is that things will move through that production system.

[00:11:28] Roberto J. Arbulu: And as they move through the production system, there is a cycle time. The time it takes to traverse. Right? Piles, drawings, right? Packages of work, deliverables in engineering, right? Even if you isolate just the engineering box, there is an engineering production system, right? Do we produce in engineering?

[00:11:55] Roberto J. Arbulu: Right? Yes, we do. Right? We produce deliverables, we produce designs. So when we, when we hear about production, it’s not just about physical production. In case we’re thinking that way, right? Is we have knowledge work, we have craft work, right? And we can see the entire system as a production, the entire process as a production system.

[00:12:20] Roberto J. Arbulu: And so one way, once again, the thinking behind what these organizations are, are applying is the idea of focusing on four verbs, five letters, and three curves, right? And this is sort of the definition of for your production management that I’m going to take you through very briefly. When we say four verbs, the idea is that because production is about transforming things, right, I transform engineering into something physical, then move to the site, is that we need to focus on the core production activities, how we design, how we make things, how we transport, how we build and commission things.

[00:13:00] Roberto J. Arbulu: Okay. So if we don’t touch this core Right production activities. We need to stop and think, what are we doing right? We don’t influence the performance of these four, production activities. We need to think about what exactly are we doing? The idea of the five levers is that if we want to influence cost time and cash.

[00:13:25] Roberto J. Arbulu: We have five things that we can actually play with from a production management perspective. One is we can do a lot of work on the product design, right? Gary mentioned modularization, right? You’re touching the, the product design. But if we modularize a facility, we’re going to have huge implications immediately.

[00:13:49] Roberto J. Arbulu: on the production process design. We’re changing the one that follows. That’s why you see a plus sign, they’re actually interconnected. Right? If you make one change in one of them, it influences the other one. Right? And the production process design, although the production work is not in there, that’s what we mean, is not really a schedule.

[00:14:10] Roberto J. Arbulu: Right? It’s the how we do the work. Right? It’s how we do the work. The third component or lever, is The idea of capacity. Okay? Capacity and resources are not synonymous. Capacity is a rate, right? And we have some resources that are contributors to the capacity of a system. Engineers and engineering, equipment, even space, right?

[00:14:40] Roberto J. Arbulu: And so, capacity is another component that is very important in a production system. And the last two, that I left purposely at the end. are typically blind spots in the way we look at the impact of a production system on, on project performance, right? Inventory and variability. When we say inventory, we don’t mean stocks of materials only.

[00:15:06] Roberto J. Arbulu: There’s another element hidden behind inventory, which is The inventory of work, the work in process, the amount of work that is either being processed or the amount of work that is waiting to be processed, right? So there are two components behind inventory here. And variability, you just did a poll on what is your sense of how things change on a daily basis.

[00:15:31] Roberto J. Arbulu: We are exposed to very variable systems, very dynamic systems. In the project world, we don’t know that, okay? And so the idea is that we can play with all these five as a means to influence cost, time, and cash. You might be thinking, so how is this different than more conventional project management, right?

[00:15:59] Roberto J. Arbulu: And so we wanted to bring this comparison to discuss that from a project management perspective, we typically see three There is that are actually a play and these come from the iron triangle mentality, right, where we can play with the scope and quality, do the schedule and the resources to achieve the same, the same outcome, right?

[00:16:22] Roberto J. Arbulu: So production management perspective, we have much more options to play with much more realistic because we’re looking at how we transform work and influence the way the work actually occurs. The idea of three curves is basically a simplification, of what we’re calling operation science. And let me start with the one in the middle quickly, right?

[00:16:48] Roberto J. Arbulu: So in any production system, if the utilization of the capacity increases to 100%, right, get closer to 100%, we will quickly see that the cycle times will start going up, right? We’ll start increasing, right? So, but unfortunately, we tend to try to utilize our capacity always to the maximum, right? That has a very negative effect on, on performance.

[00:17:15] Roberto J. Arbulu: If you look at the one on the left, there are two charts in here, but basically, if we have in a system a lot of working process, excessive working process, it gets to a point where the throughput doesn’t change, right? And we have A lot of work not finished, but open, right? When I was younger, when I was a superintendent, building things, one of the things that I was really good at is at opening work fronts.

[00:17:46] Roberto J. Arbulu: I was a work front opener like you have no idea, but I was not a good work front closer. I struggled, right? I just didn’t know all this stuff back then, unfortunately. And so I did not really understand the impact off working process. And when I say work from is obviously a construction example, but in engineering could be the amount of the liberals that we keep open right at any given time in a system, right?

[00:18:19] Roberto J. Arbulu: The more we put in that system, what this is telling us is that our cycle time now we’re looking at the chart to read our cycle time is gonna go up. Very rapidly. So if we’re trying to be more predictable, if we’re trying to compress the schedules, right? One of the things that we need to keep in mind is our ability to reduce the work in process in any production system, physical or non physical.

[00:18:48] Roberto J. Arbulu: Okay. And by the way, that’s not my opinion. I’m not giving you my opinion. I’m describing the elements of operation science. Right? And it is going to happen that way on on projects. I’m not going to talk about the other one just to move forward. But this is what we see every time that we ask questions, for people like all of us in the industry, right?

[00:19:14] Roberto J. Arbulu: To draw what is the relationship between cycle time and utilization and working process? And we get this sort of noodle soups, right? We are not in agreement. As members of the same team, right? As, companies in the industry, right? If you look at this, cycle time, let me see if I have a laser here, yep. If you look at the one below one coming down, this team, the individuals in the team that will believe this is going to happen, they will increase capacity utilization to move faster.

[00:19:53] Roberto J. Arbulu: The other half of the team, Will do exactly, well, the opposite, right? So they will try to, they understand that if they increase capacity utilization, things will actually take longer. Right? And therefore they will try to maintain some level of capacity utilization. But those are two opposite strategies.

[00:20:14] Roberto J. Arbulu: And we see these in members of the same teams.

[00:20:21] Roberto J. Arbulu: Is this making sense? Right? And so one of the things that the Project Production Institute is doing. Including this impulsion is to start creating much more than just awareness, right? Understanding the impact of these decisions on our ability to perform and perform effectively and get our projects through the process, right?

[00:20:45] Roberto J. Arbulu: It’s with

[00:20:46] Chet Carlson: mapping up the process and there are tools out there that you can use to do so. And once you’ve mapped the process, there’s specific information you have to collect that will

[00:20:57] James Choo: tell

[00:20:58] Chet Carlson: you how the production process is going to behave. So we need to know what is the sequence of operations, who does it, what policies are in place to control production.

[00:21:08] Chet Carlson: And once you have that data in there, you can start to generate useful insights for your projects. So here’s an example of an engineered production system. Their goal was to Create 1, 700 engineering deliverables. In this case, they were isometric drawings. Between May of 2020 and October of 2020. At the time, they were considering reducing the amount of people dedicated to this project and putting on another project.

[00:21:39] Chet Carlson: But they were struggling with a lot of rework. And so what you can see here is this production process that we mapped. And these three streams of work are in the same sequence of operations, but one with partial rework. One was a more complete rework of the, of the ISOs. And they were having difficulty computing how much of their capacity was eaten away at because of the rework.

[00:22:02] Chet Carlson: So we created this model to tell them and show them exactly what’s going to happen if they keep on track with their power plant. There was a team that was installing a massive steel structure on site. And the question arose, struggling a lot with the capacity of their cranes. Because they were lifting each individual component one by one.

[00:22:27] Chet Carlson: And so, they weren’t going to make it because there were just too few lifts and they were constrained on the number of cranes to fit in that area. So what we, what we took a look at is we estimated a couple different ways of pre assembly. One without pre assembly, and you can see they totally blew that.

[00:22:41] Chet Carlson: The second one, 30 percent pre assembly. And the little bottom one, 45 percent pre assembly. So, most, almost half the work needed to be pre assembled before it was installed in order to keep up with their current team.

[00:22:55] James Choo: And, I’m going to

[00:22:59] Chet Carlson: When we’re modeling production systems, we have two ways of analyzing how the production system performs. You’ve got analytical modeling and discrete simulation. We’re going to go over a couple of examples each in just a moment. But at a high level, the analytical modeling allows us to find optimal policies, sort of design the production system, so that if those policies are in place, you get the results that you want in your project.

[00:23:27] Chet Carlson: And the discrete event simulation allows us to track how this project, how the production system is going to behave over time. And we can see whether or not specific deadlines are met. You can see the accumulation of work in process in various parts of the project. So you can plan for the space required.

[00:23:44] Chet Carlson: This first project was a 14 column recurve array. And we applied, decided to use analytical modeling as a process, it was quite simple. Essentially what we’re doing is, we’re driving precast piles, sheet piles, here peer and mean, and columns, and slabs. And so it’s a very simple process here at the top.

[00:24:07] Chet Carlson: And what this model allowed us to uncover was, in each chunk of this process, what’s the bottleneck? Because that’s really what’s driving the rate at which you can go, how fast you can go. And, if each chunk, is there an opportunity to change some policies to go faster? And how much work and process should we have built up.

[00:24:25] Chet Carlson: Time, which you can see from this previous picture, is they have a ton of work that’s started, but a lot of work that’s finished. What the team reported after implementing some of the strategies that we gave to them was a 300 percent improvement cycle. And this was actually twice as fast as the classical B applications that they were tinkering around.

[00:24:49] Chet Carlson: So it was a huge gain to that. The way we did this was we took, we printed a model for each individual. Workflow in that process. So you’ve got precast piloting, sheet piloting, poly cap construction, column construction, peer construction. And in each of these different workflows, we identified potential opportunities to get the schedule schedule savings on, right?

[00:25:14] Chet Carlson: So you each, most of these, it was around bottlenecks, adjusting bottlenecks. And in these bottom three alcap construction column construction area construction, there was significant handover sides between. Each team. And so what that was effectively doing was increasing the huge batch and pushing up their cycle times.

[00:25:36] Chet Carlson: Now for each of the bottlenecks, each of these bottlenecks contributed a potential delay to each to the schedule here. So if they didn’t address the bottlenecks, then this is what they would expect for an additional delay project. It’s about 100 days. So for each workflow, one of the critical elements to understand bottlenecks was to Map out the capacity utilization of each resource.

[00:26:07] Chet Carlson: The team knew they had to add people, but they weren’t sure about which people and which equipment to add to increase capacity. And it’s important to know this because if you want to go faster, you need to know where you need to decrease capacity. And ultimately what drives the speed at which your project can can dredge through.

[00:26:29] Chet Carlson: So if you’re adding resources, you don’t want to add people here. Where utilization is basically, it’s got to understand your out, start to fall off, but one change in a parameter is going to change this, this, this profile completely. So it’s important to understand when you bring people on and when you’re taking people off.

[00:26:46] Chet Carlson: When is the best time to do that? When should you be manning up? When should you be taking off? We can help them. Identify when is the best time to use and bring on additional

[00:26:54] Roberto J. Arbulu: capacity. If you go back to the previous one, I just want to highlight something for it. So you might have noticed that these production system models have these little boxes, little triangles.

[00:27:04] Roberto J. Arbulu: And so you might be wondering, where are those triangles? So the boxes are actually where the work actually starts to form. And the triangles is when the work actually waits. Meaning nothing is being done to it. Right? And if you start adding the time, because there’s a time component in these little triangles, right?

[00:27:24] Roberto J. Arbulu: You start adding all that time across the, the, the production system, right? We can measure by mo and we can actively measure, quantify, right? Not a guess quantifying what happened with the total cycle time and what we’re learning as, as we do this and this, and this allows the, the, the, the world actually is that the opportunity.

[00:27:46] Roberto J. Arbulu: To compress the cycle time is not so much on the operations themselves is avoiding that the work waits for available capacity, right? And is the time is in this is in the strength. And so we need to shift our thinking from productivity to the productivity of the system, which we call a throughput. They are not the same.

[00:28:16] Roberto J. Arbulu: Okay. Remember, we’re not doing a P50, P90 schedule analysis or anything like that. We’re not analyzing the right side for that schema, we’re analyzing the left. So there was a question

[00:28:31] James Choo: at NBS you still, it trams, I mean, not all work is created equally as it goes through the system. So if you have a value of work that sits behind at the time you, is there a way to rather relatively show the damage that work should in your versus work should later in the system?

[00:28:50] James Choo: Because it’s not all, it’s, I mean, depending on where you are in the project, it’s not all created equal with, in terms of the impact. So that makes sense.

[00:28:59] Chet Carlson: If work sits at the end of what? Well, not all

[00:29:03] James Choo: work is created equally as it ever should in the system. The longer the work sits at a certain point, the time it would be ignored.

[00:29:09] James Choo: The time and expense where it sits might be less than the value of the work sitting behind it. So I’m just trying to, how do you maintain that as you look at the overall effectiveness of the system?

[00:29:20] Chet Carlson: Yeah, so typically what we’ll do is we’ll try to understand, you know, at some point all the work is connected, right?

[00:29:27] Chet Carlson: So we build that into the model.

[00:29:29] James Choo: Oh. Now, a work is created before it has ever existed. The longer, the longer the work sits in the circle, like the time that, or the time and expense where it sits, might be less value to work sitting on. So, let me try, how do you

[00:29:44] Roberto J. Arbulu: Maintain that as a good people.

[00:29:48] Chet Carlson: Yes, a good people will do is we’re trying to understand, you know, at some point all the work is connected, right?

[00:29:55] Chet Carlson: So we build that into the model, how work is connected and how things hold true. And so certain streams of work will have to happen, right? And you’re going to sit and wait for it, last one, and then they get pulled through from there. Then it ripples on down to the end of the system. And so we’re able to capture what are those relationships in between the work and pull it through

[00:30:13] James Choo: the system, right?

[00:30:15] James Choo: But it is driven by capacity utilization. So as the work is in, so that’s why we call it the second wrap, where it’s actually a combination of variability of the work to the end, where all of this work is in. And how much variation is there in the word cell, but it is driven by the actual visualization does the work.

[00:30:39] James Choo: So that’s why you can’t really just actually have a single item going through the system and actually determine what the Qt app will be,

[00:30:48] James Choo: but it did.

[00:30:52] James Choo: So

[00:30:54] Chet Carlson: just one more note, here’s the active variability. So what we’re plotting here, we ran a simulation, I don’t know, maybe a hundred times. And you can see because of variability, because the policies that are chosen for a particular project, there is a distribution of, at, this is the completion of the project.

[00:31:13] Chet Carlson: So the, the time that the final task is finished. And what we want to do is, is we test different strategies to see how can we tighten up this curve, and then how can we bring it to a point where everything fits within the portion or within before the milestone target.

[00:31:34] Chet Carlson: Ah, and then just to recap, so this project team, With the two, modeling applications that we we gave to them, they were able to achieve 20 percent scheduled accelerations, 300.

[00:31:51] Chet Carlson: And if you want to read more about this project There’s a paper that was just published and it adds a lot more detail. And it’s the Thompson

[00:31:58] James Choo: Instructional User Accounting website.

[00:32:02] Roberto J. Arbulu: Yeah. So, by the way, we need to acknowledge that this is a piece of work that we did in partnership with our friends from McKinsey and on some effort that we collaborated with and, we, we got an invitation by Kurt to talk about it.

[00:32:23] Roberto J. Arbulu: So, take a look at that. And that article will occur well, if you were interested in getting more details on the story. So the last piece is, as we’re trying to answer this question, what is going to happen on my project? We can do the volume, but as we’re executing work, we know that we’re going variability.

[00:32:48] Roberto J. Arbulu: Right. So what are teams doing in order to control that and also make sure that work in process that will not actually get out of control. Right. And it’s a second strategy, which we call it production control that is coming up. When we talk about production control, we don’t mean great controls, by the way.

[00:33:08] Roberto J. Arbulu: OK. Controls is not the plural of control. OK. I repeat, controls is not the plural of control. Right. Controls are more focusing. Getting progress control is more about direction, right? Control is more second direction. So I don’t need to spend a lot of time on controls, but, you know, these are examples of what people are doing.

[00:33:34] Roberto J. Arbulu: They use a baseline schedule, rules of credit, and we need their value, right? That’s typical early controls. That’s not what we’re talking about here, right? We’re talking about the ability to control the capacity. Inventory and the vulnerability as we go through and we need to put different means and methods of how to do that.

[00:33:53] Roberto J. Arbulu: But that’s what you see in these photos to be able to influence is an example of engineer engineers understanding the process design. And this is a construction team understanding as well. They’re on process design as we, as we go through. So we’re going to expand on this. So this engine doesn’t replace the current efforts on scheduling.

[00:34:17] Roberto J. Arbulu: But it actually augments because it is a production control engine. And this production control engine has four main components. The first one is relentlessly focusing on standardizing production processes. Why do we care about standardization of production processes? Because we can actually compress.

[00:34:38] Roberto J. Arbulu: Cycle, right? Not only standardizing the product, but also the other standardizing process. Second component is combining those standard processes to create an integrated production schedule that might look very colorful. Because one of the things that we’re Deeper understand is how the work is flown and how it changes forms,

[00:35:05] Roberto J. Arbulu: because we know that what happens in between interfaces, there’s high likelihood that the work will wait, that we will have queues. And if we have a lot of work reading, the cycle time was again, get extended. And so, and also it’s understanding the true complexity of the production schedule. Then we do a lot of work on production planning.

[00:35:26] Roberto J. Arbulu: We make decisions about how we allocate the limited resources that we have, the detail on on the work and then get it. A set of non traditional analytics is a lot about risk curves and CBI, STI disease other type of analytics that will uplift, right, their perspective. It’s almost like giving the team a different set of levels to be able to see new opportunities or how they could actually improve their work, right?

[00:35:55] Roberto J. Arbulu: This is one output is an oil and gas company that executed on the same project. Some piece of work without production control and another piece of work with production control. You can see just graphically, without reading anything, that the one without production control had less, certainty, right?

[00:36:19] Roberto J. Arbulu: Bell curve is much wider, and the actual output with production control, much more predictable. So they were able to increase their predictability to about 60%. Why it’s not 100?

[00:36:38] Roberto J. Arbulu: We’re dealing with variability, right? Right? And so, remember, we’re predicting. If we are able to predict only 60 percent of the world tomorrow, right? Look at what they achieved. First gas milestone in trade 2, with one day of the forecast that was established 9 months earlier, right? This is because what they implemented is that engine.

[00:36:58] Roberto J. Arbulu: They ended out non stopping, week in, week out, right? To be able to control the variability. This, this is our only example in engineering reviews. A good amount of pos The connection between India, and gen is very tight. And, they’re saying these took great deal of coordination and production workload, instrumentally made it happen.

[00:37:27] Roberto J. Arbulu: they were actually started to late, as you can see, the will, serve and they finish about 14, 14 days ahead of the schedule by putting, I repeat this portion, this element, augmenting, complementing. What they were doing with the schedules and registers. Okay, this is Terminal 5 in London two years back.

[00:37:51] Roberto J. Arbulu: Still, a lot of people talk about Terminal 5, but they did something exactly that, and they were able to finish the project, with a significant amount of savings and time. It was the first UK civil project to deliver on time in the last 40 years, but they applied all these techniques. This is another project in Australia.

[00:38:15] Roberto J. Arbulu: It’s a

[00:38:16] James Choo: desalination plant in

[00:38:18] Roberto J. Arbulu: Australia that the team concentrated quite a lot on the way they were producing during the construction the work, as you can see in this photo, understanding where the variability was coming from, looking at the sequence. Making sure that they don’t open unnecessary work fronts.

[00:38:35] Roberto J. Arbulu: They don’t create working processes. And also start measuring, right? How they were moving over time. In terms of becoming more reliable. And able to achieve the results. 15, 18 days ahead of their schedule. And so, we finish with

[00:38:52] James Choo: this. All we’re saying is.

[00:38:55] Roberto J. Arbulu: If we follow conventional thinking. The likelihood.

[00:39:01] Roberto J. Arbulu: That our production ability is extremely high, as you saw in those statistics. We need to do something radically different to look at this. And the idea we bring to you today, think about it, is focus on the production system. Because if the production system doesn’t have the capacity to deliver what we want, we are not going to make it.

[00:39:23] Roberto J. Arbulu: More scheduling is not going to help us. More digital scheduling. Help us. Any questions to me?

[00:39:35] James Choo: Yeah. You know. Great, great. One thing I wanna ask while, while the customer, I wanna go back to your , right? You we’re talking about ness ion start, predictability of that. So by thought. Why are they going to fit for a 5G network?

[00:39:58] James Choo: Or something if, yeah, they don’t satisfy the preferences? So, well, you see, in average, they are, these are critical as far as RAC2. So, I’m just wondering your thoughts. No,

[00:40:10] Roberto J. Arbulu: I, I agree with you. We can, we can do that. But when we say build, this is where we are, through the commissioning, that you’re, you’re making.

[00:40:19] Roberto J. Arbulu: In commissioning, we, the production system produces systems and subsystems ready to be operated, right? And DPD in industrial facilities, or even in other ones, we see the transition between construction to commissioning, the transition between working based on areas to work based on systems. And what commissioning needs, because of a specific systems and a specific sequence, is not necessarily what construction releases to that.

[00:40:46] Roberto J. Arbulu: Right? And so the commissioning teams spend a lot of time defining the optimal sequence of commissioning, but what, what construction give them is a different kind of work and things get completely sideways. And so what we are learning in terms of commissioning is we can model, we have activity, we don’t have an example here show, but we have model commissioning production systems and they link to the transition from areas to system based execution.

[00:41:13] Roberto J. Arbulu: It’s a very interesting thing and there’s a lot of patching happening there, by the way, right? Because for a specific, specific system, you might need multiple subsystems. If you are missing one subsystem, the commissioning on the system waits. And the batching create a lot of weight. Yes, sir.

[00:41:30] James Choo: And

[00:41:39] Roberto J. Arbulu: microphone. Use this one.

[00:41:45] James Choo: So the actual question I had was in a commissioning, you typically have like, you know, 300 pieces of equipment in a specific data center. Each of those have an, has an OEM that has their own unique supply chain, right? So If any one of those pieces of equipment in the commissioning is late, you cannot really commission the entire system.

[00:42:08] James Choo: So I really get the concept of project control and how that’s different from project controls. That’s, that’s awesome. My specific question is how do you model the entire system? You’d have to go to each OEM supply chain in order to fully understand. The constraints of the bottlenecks of each supply chain, and I’m talking about companies like Schneider, for example, that has very complex global supply chains.

[00:42:39] James Choo: So how do we model the system while we’re trying to understand, you know, each complexity of each supply chain at the end of the day, as an owner company at Microsoft, everything needs to come together. In order to commission it, right? So, just wondering how detailed you go into the project control modeling for that.

[00:43:00] Roberto J. Arbulu: I have a surprise for you. Ready? What you see here is one of those production systems of one of those companies that you mentioned. This particular order, right, in the data center space. Decided to get a bit deeper on one of the multiple equipment they supplied, right, and analyze it because they smelled something wasn’t actually brought to be okay, as they were planning to increase the capacity that would be fixed.

[00:43:35] Roberto J. Arbulu: And so we can include a lot of supply flows. You’re going to see, by the way, this exact, very interesting schematic from Brian Green from BITCONNECT later today. So keep an eye on what he’s going to bring on, on exactly your question. But this is an example of a module that is fabricated, not in the United States, it’s in Europe, okay, more specifically in Ireland and doing that.

[00:44:08] Roberto J. Arbulu: So we got to dignify these. Three detail one, but we can simplify it, connect the entire supply chain and all without any too complicated, right? And sometimes, honestly, in our experience and we see things trying to simplify complexity versus modeling the complexity, right? If it is complex, it is complex.

[00:44:34] Roberto J. Arbulu: So be it, right? Let’s analyze it, right? And let’s, let’s bring what’s going to be the impact of that to a complex.

[00:44:44] James Choo: So you do this at a high level for each of the various, like the supply chain. So you could do it up at a high level. Okay. Great. What a perfect sense. So, you know, I really, really like the application of this on the construction industry.

[00:45:00] James Choo: This is very common to the manufacturing retail industry, probably 20 30 years ago. Value Street Map, the Little’s Law, you know, that’s very, very common there. You know, the, you know, S& OP planning, demand supply planning, that’s very, very common there. But I really like the application of the capital projects construction industry, so I agree with you.

[00:45:23] James Choo: I think there’s a lot of but we do have a lot of work to do as you guys have said, we’ve been on this for 10 years now, right? So we definitely have a lot of work to

[00:45:33] Roberto J. Arbulu: do on this. That’s what we’re meeting today. Yes, sir.

[00:45:44] James Choo: So if you’re working with the existing project that has large batches of work, like a distraction contractor that, It’s left of work packages at 10,000 hours versus thousand 600. You have to go back and rework the requirements that’s scheduled, like

[00:46:01] Roberto J. Arbulu: just be able to haul it. We can model because the, the size of the package, right?

[00:46:10] Roberto J. Arbulu: As an applications on, on or the batch of the work, right? You Apache, a little work or I’m gonna work, right? And we don’t model the act of what happens if the size of the package is big or small. Because you don’t see, actually, I remember Shema was actually doing an analysis where they, on two subprojects on the same project, one of the projects had larger packages than the other one, for some reason, right?

[00:46:37] Roberto J. Arbulu: This package, the package was smaller, and those guys were able to move faster, just because of the amount of work that was being released into the field, right? They were able to reduce the working process. Who have and does have, Connected. Yeah.

[00:46:55] James Choo: One question, you have the production system, And then we have the producers, And one thing we’re working on CISIS, is the culture.

[00:47:04] James Choo: Alright, how do we change that culture? And the place that I noticed it, That we got interested, was actually in T5, St. Patrick’s. I don’t know where it is. They used systems like this. But I’m wondering, to get your perspective, it. You know, on, on, on that, not only from, you saw a package of knowledge, workers, craft workers, but we need a lot of Kazakhstan and, you know, the knowledge and the craft, you know, it’s basically World War II communist versus, you know, Gulf coast.

[00:47:34] James Choo: You know, there’s, there’s certain cultures that are in there that help you attack. And we’re trying to figure this out, how we accelerate it. And of course there’s tools out there. I said, I’m going to be here.

[00:47:44] Roberto J. Arbulu: Your thoughts. So, craft work, craft workers work in the system, right? Work for the system, right?

[00:47:56] Roberto J. Arbulu: Knowledge workers,

[00:47:57] James Choo: you want to make sure that the system

[00:47:58] Roberto J. Arbulu: works for them, so to speak, right? And so if you decide the production system,

[00:48:04] James Choo: physical production system, even if you’re

[00:48:06] Roberto J. Arbulu: in a different country with a certain what we have learned is that that can be addressed by a more effective design of the actual production system to mitigate the cultural or even less productive sort of working market.

[00:48:24] Roberto J. Arbulu: And now, a lot of companies do training and things that also help, right? They try to use this,

[00:48:29] James Choo: invest in training centers and things like

[00:48:31] Roberto J. Arbulu: that, right? All we have learned is that if you focus on the production system, the cultural factors can be activated. address as well, right, because they work in the system.

[00:48:41] Roberto J. Arbulu: Now, let me go back,

[00:48:43] James Choo: because we found, whether we were in China, we were in Kazakhstan,

[00:48:48] Roberto J. Arbulu: Russia, Korea, we were in Australia, or the U. S.

[00:48:53] James Choo: When this got down to the people doing the work, it

[00:48:56] Roberto J. Arbulu: provided clear direction, clear understanding,

[00:48:59] James Choo: less rework, less confusion, and zero

[00:49:03] Roberto J. Arbulu: for the

[00:49:03] James Choo: people. The cultural challenge was People between there, who said, I don’t believe in this stuff, I want to do what I want to do.

[00:49:14] James Choo: That became a harder challenge of it, to, you know,

[00:49:17] Roberto J. Arbulu: break the rules, you know.

[00:49:21] James Choo: Yeah, well, I, you had something to add too? I think for now you actually ended on the word, technology, right? And if we actually look at, let’s look back maybe 12 years. Remember the days we got cell phones? Do we actually change our mindset when we adopted cell phones?

[00:49:40] James Choo: Do we actually have to get an IT department all signed off before we all started walking around cell phones? I remember the days when big corporations did not allow employees to have an iPhone because they couldn’t control it. Remember? And then after a few years, because of the mass growth in the use of the iPhone, IT companies had to figure out, IT departments had to figure out how to incorporate it into their system.

[00:50:02] James Choo: So we think technology has a huge role to play. The problem that we see is right now there’s a lot of automation. Of the existing methodology going on, even if it’s optimized, it’s reducing maybe your RSIs rather than actually expanding it and rotating it from all aspects. So I think we have a huge opportunity right from the start where we can actually use a different framework to generate a new level of digitalization to change how we do it.

[00:50:37] James Choo: Many, many people work in the system without even knowing that they are a part of the system. Right. So when we, so we actually have this, this is an Apple education training program for manufacturing industries. Believe it or not philosophy for a month. I’m doing this day. And one of the things that one of the person that she says, I learned more about the production system, attending a five hour course, the idea of working seven years, because they were part of the system, they came to work, they operate.

[00:51:06] James Choo: It is some of the rules and governance that was provided by the company we operate in New Zealand. So I think what we’re saying here is, let’s design the system properly, so that it can do a lot of other things. One of the aspects that Veeam talks about is respect for people. So let’s give them an opportunity to see a pretty good writing environment.

[00:51:27] Roberto J. Arbulu: One quick, now that you mentioned that, Greg, we’ve got our BBI, but we saw the other day an application of robotics in construction, which was the use of robots. To track progress, right? There’s nothing wrong about that, as long as you understand that the technology is being used to feed the Quake administration machine.

[00:51:55] Roberto J. Arbulu: The use of that technology will have zero impact on performance outputs. Zero, right? Because it all just goes through and tracks progress. Great, they’re using it, but if the connection is made between the use of that technology and their performance, then we have a problem. Because it doesn’t impact anything.

[00:52:15] Roberto J. Arbulu: Maybe it reduces the cost of a couple of guys walking around digging, maybe. Right, but how much? Well, if

[00:52:23] James Choo: they use it, like, to validate completion of the

[00:52:28] Chet Carlson: task.

[00:52:29] James Choo: Oh, yeah. I’ll tell you, you better snap you up,

[00:52:33] Roberto J. Arbulu: because work is out of it. That’s a different thinking. Correct.