Featured Presentation

Project of the Future – Vision

Exploring the Integration of Human and Digital Capabilities in Modern Construction: Insights from Martin Fisher, Professor at Stanford University and Director of the Center for Integrated Facility Engineering

Join PPI

Overview

Martin Fisher, a professor of Civil and Environmental Engineering and Computer Science at Stanford University, as well as the director of the Center for Integrated Facility Engineering, a senior fellow of Precor Institute for Energy and coordinator of the Building Energy Efficiency Research at the Preco Energy Efficiency Center, gave a presentation on the topic of blending human and digital capabilities in the construction industry. He discussed the concept of “modern construction,” which brings together three main trends in the industry: industrialization, autonomy, and digitalization. He explained that industrialization focuses on standardizing products, organization, and processes, while autonomy refers to machines taking over certain tasks and decisions. Digitalization enables autonomy and improves decision making in project production. He also emphasized the importance of this topic due to the increasing expectations for facilities to have strong economic, environmental, and social performance. He discussed the challenges that companies face in meeting these expectations and how blending human and digital capabilities can help to address these challenges.

Transcript

[00:00:01] Gary Fischer, PE: Well, as promised, this is one of my favorite parts of the day that I’ve been looking forward to all day long. Welcome, Martin. This is Martin Fisher. He’s the professor of Civil and Environmental Engineering and Computer Science at Stanford University. He’s also the director of the Center for Integrated Facility Engineering, a senior fellow of Precor Institute for Energy and coordinator of the Building Energy Efficiency Research at the Preco Energy Efficiency Center.

[00:00:26] Gary Fischer, PE: His research focuses on modeling, predicting, measuring, and improving lifecycle performance in the built environment and is a good partner with PPI. So Martin, we’re not going to pinch you on your time. You have as much time as we allot. We’ll take it out of the panel discussion afterwards. So welcome and thank you for your efforts to pull this together.

[00:00:43] Gary Fischer, PE: Look forward to what you have to share.

[00:00:46] Martin Fischer, PhD: So no, thank you for having me again. Always great to participate in the symposium in this community. And it was great to see the work of KMO and, as you could see there already, it’s a really excellent illustration of really the main point I would like to make, which is that, in the future, well, I mean some ways already today but, in the future, much, much more so the production systems we use on our projects to produce the digital work, to produce the physical work onsite, offsite in the office is going to have.

[00:01:25] Martin Fischer, PhD: Combine human and digital capabilities much, much more strongly than we see today. They might say, well, come on, Martin, you’re already using lots of computers everywhere. But the opportunity we have now, I think, as you can also see in your work, is to connect the different discipline phase silos much more through visualization data integration.

[00:01:53] Martin Fischer, PhD: And automation and to some extent in some cases even autonomy of what the machines do. And that is quite a new world that we are really entering. I would say so far we have been trying stuff. If I sort of extrapolate and put myself 10 years out, I think so far we have been experimenting.

[00:02:13] Martin Fischer, PhD: I think people look back in 10 years and say, yeah, those were some cute applications we had. Now we are working at a totally different level and I think it’s really this combination of human and digital capabilities that excites me. And we’ve coined this phrase together with PPI; we call this “modern construction.”

[00:02:34] Martin Fischer, PhD: So the key question then is, “Why is blending and digital capabilities so important?” What are particularly interesting digital technologies? How will the roles be of humans and digital technologies? And then what are the strategic and operational management challenges for a production system and for a company that uses such production systems that blends human and digital capabilities seamlessly.

[00:02:57] Martin Fischer, PhD: I’ll try to shed some light on these questions in my presentation. I mean, that’s also a whole class but hopefully enough to stimulate the panel discussion. So modern construction, I mentioned that the term really brings together three main most impactful trends that I, well, actually I think we see in our industry industrialization really thinking about what is bespoke on a project and what can be.

[00:03:28] Martin Fischer, PhD: Standardized in terms of product, organization, process, what can be done off site. Onsite, but it’s beyond just onsite, offsite. That’s a big part of it. It relates also to the business model that is possible, enabled, required. So that’s a really big trend we see everywhere in the world, in all sectors of the industry.

[00:03:49] Martin Fischer, PhD: Autonomy. We see that, increasingly, that machines take agency – take things off our hands. And of course there’s also many challenges still with that. Self-driving cars might still take a little while to be fully self-driving. But in building or factory automation, we see, certainly, certain decisions being made just by the system.

[00:04:19] Martin Fischer, PhD: Based on data and then the whole field of digitalization, which really enables autonomy in this industrialization. But also addresses other aspects of project production in terms of decision making, for example, that project teams have to make. But I’ll hopefully illustrate these elements a bit more and be able to bring them together through well illustrations and then combination of those illustrations.

[00:04:47] Martin Fischer, PhD: But why is this so critical today? And just to remind us, I think there’s significant challenges here. I don’t really recall a period in my professional life when we’ve faced these kinds of challenges at this scale. I mean, for example, the expectations for our facilities in terms of their economic, environmental, social performance.

[00:05:10] Martin Fischer, PhD: Just keep going. Because we always had the economic pressures, but now we need to also show great environmental and social performance, which rightfully so. But we don’t always, most companies that I work with don’t really have the project production systems in place to truly consider all of these.

[00:05:34] Martin Fischer, PhD: Then huge demographic shifts in terms of across countries, but also when in, if you just look in industrialized nations, our customers are changing who the customer is. They’re getting older, for example, etcetera. And engineers, builders are getting younger if you can attract them to your company.

[00:05:57] Martin Fischer, PhD: I mean, that’s what I hear around the world is just the challenge of getting a skilled workforce. So we have to come up with a very different way of making decisions on projects. You cannot rely, in a few years, I think maybe already today, on people that have had 20, 30 years.

[00:06:20] Martin Fischer, PhD: We need to find a system that allows younger people with less experience to still make good decisions. And that’s certainly possible. I’ve seen a dramatic increase in learning when we bring digital methods into the mix. But that’s another reason why I think we need to find a blended production system between human and digital capabilities and then resource availability.

[00:06:47] Martin Fischer, PhD: In terms of materials, yeah, labor already mentioned. In some cases also equipment. So I’m sure there’s more challenges, but these by themselves are already, I think, big enough challenges. But I think the good news is also that we have dramatically greater possibilities today than again, if I think back on my career, we’ve had.

[00:07:05] Martin Fischer, PhD: In terms of the three topics I already mentioned, industrialization, right. Raises the question and the opportunity in terms of where do we produce so we can have good productivity, safety, quality, reliability and can have better resource use. And that relates back to the successful companies, they’re blending that.

[00:07:27] Martin Fischer, PhD: Of course with changes in business model in terms of what it is that the company actually guarantees or what it actually – how much of the value changes – it’s actually covering. And it’s interesting that I’m seeing kind of different strategies. So I’m focusing on the part of the production, the overall production process, and not of saying no.

[00:07:47] Martin Fischer, PhD: Now we have the opportunity to really control everything and I see successful companies with both of those. But this is a decision now that in engineering construction companies, you have to make, because if you’re an engineer and you have to compete with somebody that is offering an industrialized product, well the engineering is included there.

[00:08:07] Martin Fischer, PhD: So how much can AC engineering cost when you are in the market segment where it comes for free with what somebody else is offering? Just as an illustration? And then the autonomy really advances in basically control systems and with communication advancements and computer vision, for example, to create these rapid feedback loops that allow autonomy and continuous learning.

[00:08:34] Martin Fischer, PhD: I mean, that’ll continue to evolve, but in some parts of construction, we can already see autonomous devices deployed, as you will see. And then the whole world of digital with many, many technologies for mobile to cloud optimization. The interaction between physical and virtual in around location dimensional control, machine learning, AI, robotics, additive manufacturing, Internet of things, and sensing more broadly virtual reality, augmented reality, extended reality.

[00:09:10] Martin Fischer, PhD: These are some key technologies that each one on their own already is a big deal, but actually coming into our world with innovations almost every day, too. So I think you can see that we are looking at, at least that’s what it seems to me, quite a different world in terms of the challenges we have to address but also by the means we have to address them.

[00:09:39] Martin Fischer, PhD: And this is not really something that we necessarily actually, in academia, know how to do. And industry, I’m not sure there’s a company that actually is really on top of managing all of this. And I think this is where an organization like PPI, in my mind, is so important to bring us together and develop how we tackle these challenges and leverage these sorts of possibilities.

[00:10:11] Martin Fischer, PhD: So why is this a big deal? You know, I mean, we are building projects today, but consider a few performance criteria here. Cost, right? Digital tools that cost you cents per hour versus many, many dollars per hour. And I understand an hour on a computer is not the same as an hour in person, but when…

[00:10:35] Martin Fischer, PhD: …you have a person do something that could be done by a computer, that is a problem. Looking at just the cost difference speed. We typically see when you digitize something. Ten times or even more, a hundred times faster, execution. Opening up a whole world of possibilities in terms of, again, feedback loops, exploring options and so on.

[00:10:59] Martin Fischer, PhD: Consistency. I don’t know who else other than through the means of a mix of industrialization, digitalization and autonomy. We will get to something close to Six Sigma. Well even, Two Sigma. Let’s say performance on our projects. That would already be, I think, quite amazing if you got the Two Sigma.

[00:11:20] Martin Fischer, PhD: But I don’t know how else to create highly consistent, reliable performance. And the same with scalability. Purely human practices scale very slowly and imperfectly. I think we all know that. And when we couple digital and human capabilities, we have a much better chance of actually scaling good practice.

[00:11:44] Martin Fischer, PhD: We already saw from the challenges we have to increase the scope of what we need to consider to make a good project. And again, that requires a new tool set. That has to be data driven because it’ll take us each project manager, I mean, decades to just, you know, take us decades to learn experience, to learn about the cost and schedule, performance and quality and safety performance.

[00:12:08] Martin Fischer, PhD: But now we need to add the user perspective, our mental perspective, and many other perspectives. And we, we simply cannot afford decades of building that experience. Pacing decisions on our experiences. We need to create more systematic data-driven ways of learning optimization. We have seen when we have been able to apply optimization methods that often a 10% to 50% improvement and learning I already touched on.

[00:12:40] Martin Fischer, PhD: Of course we have our beliefs and our experiences, and I’m not saying we should not use them, but when we can couple them with insights we gain from truly analyzing what happened on a project or what could happen on a project, we are much better off. And everybody talks about continuous improvement, but in few places I have seen really a sound basis for that continuous improvement.

[00:13:01] Martin Fischer, PhD: And that’s really what modern construction gives you, because it, in a way, forces you to establish a much more clear baseline and strategic vision of where you want to go. And then decision making, I think I already kind of mentioned. Do we just rely on experience or experience in facts? And I would rather rely on experience in facts, but then we have to create them.

[00:13:30] Martin Fischer, PhD: That would actually be a good metric on the project. How many decisions have we made just on the basis of experience and how many on the basis of facts?

[00:13:41] Martin Fischer, PhD: Oh, sorry. Yeah, so let me give some examples of different digital technologies. I won’t be able to do the many digital technologies really justice. But just to illustrate that, and then I’ll share a framework that hopefully is helpful for you to manage this combination of digital and physical and human capabilities.

[00:14:03] Martin Fischer, PhD: So in this case, seven professionals did a design review of a digester tank for a wastewater treatment plant in San Francisco in virtual reality. And so the digital environment really helped bring the professionals together and gave them a shared experience. And led to a design. Allowed the team to produce a design that works for all the stakeholders in a much more rapid way than in a typical.

[00:14:33] Martin Fischer, PhD: You know, a handoff kind of process that we see unfortunately too often producing a cost optimal structural design required development of new concepts of how to think about structural design and analysis, and also the aggregation of the material data, cost data, but also the erection and fabrication cost data.

[00:14:59] Martin Fischer, PhD: So all these costs could be considered from the very beginning of a project. And when we applied this to a project, much more complex than the one shown in the picture, but for non-disclosure agreement reasons, we can’t show the actual project. But when we applied that to an actual project, we saw that the integrated and optimized way of designing.

[00:15:21] Martin Fischer, PhD: Concurrent way of designing this – producing this structural design led to a structural design that was 10% more cost effective. And you can see that pretty much all of the savings are coming from the details. And when you, again, contrast this with a sequential process, you will probably not have access to those savings that come from the details because you’ve probably locked in the overall sort.

[00:15:49] Martin Fischer, PhD: Big stuff that needs to be – that makes up the structural system in the early design stages. And you might have been lucky and picked exactly the right components that lead to an efficient production process. But you might not have, in this particular project, they were not totally lucky and there was an opportunity to save 10% basically in the details.

[00:16:11] Martin Fischer, PhD: So by making the concurrent, the detailed and the conceptual design. Concurrent and analyzing many, many options. We can produce a cost optimal structural design and we’re producing the best construction strategy. Similarly the company, in this case, Clark Pacific, wanted to to know from the beginning of the project what is the “best” construction strategy and best, I put it in quotation marks because, you know, of course it’s duration, cost, idle time, resource utilization, consistency of that and so on.

[00:16:47] Martin Fischer, PhD: And so there’s many factors that play into what they consider best. But again, similarly to the structural design. It was a matter of very rapidly iterating and consistently iterating on these construction strategies and trying different ideas. So that, over the course of three days, these two professionals could come up with the best construction strategy for this project.

[00:17:14] Martin Fischer, PhD: In terms of producing a reliable construction process, a technology that has been out now for a couple years, a few years, is versatile, but there’s other methods that allow you to very rapidly and consistently document what goes on on a construction project. And here you learn what the crane is doing every second.

[00:17:36] Martin Fischer, PhD: And this allows you to really understand the variability in installation cycle times here, for example, for the Shia walls on the parking structure. And it’s actually quite surprising to me to see a variability of 23 and a half minutes for the fastest and the slowest 111 minutes. So many of you have probably placed the guys dice game where you sometimes may wonder like, what, one to six?

[00:17:58] Martin Fischer, PhD: That’s a huge variability. Well, but that’s actually what we see here in practice. But learning, right? What the variability actually is is obviously a key part of creating a much more reliable production process, either through reducing the variability or finding the appropriate base of buffering.

[00:18:21] Martin Fischer, PhD: Or building exactly what was designed one way, or this is for solar form, a fully autonomous layout of where the foundations for the different solar panels go. And so this throne flies around and basically does all of the layout automatically. And then when we compared this with the traditional layout method, we saw significantly proved accuracy and dramatic time and cost.

[00:18:56] Martin Fischer, PhD: In terms of actually producing work, sorry about tying a robot for bridges that is able to build, yeah, tie 1,200 ties per hour. That’s slightly more than I did when I was a rebel layer. And again seeing a significant improvement in safety, in particular ergonomics of the work, a reduction in labor material.

[00:19:20] Martin Fischer, PhD: And yeah, still quite a significant schedule in cost reduction. So these are just very quick vignettes of some of the digital technologies that I mentioned at the beginning. And we are using NBC around the world, everybody using, you know, several of them. And I think now the challenge is to scale that use, but also then to really put all of this combination of digital and…

[00:19:48] Martin Fischer, PhD: …digital and human capabilities together. And to do that we reflected, and I should have highlighted on each slide, but the credits are there – the work of the students that help do that work. And together with Ashvin Al able to put a framework together that at least I found quite helpful in thinking about this.

[00:20:11] Martin Fischer, PhD: Because if you think about any kind of production process, you need to observe in some way. You need to typically analyze and predict a few things. You need to make decisions, and then you need to execute actions. And so in terms of then we need to decide who does that. And for what kind of task?

[00:20:34] Martin Fischer, PhD: Who does that? So we could imagine at one end of the spectrum, a fully autonomous automated way of doing things. The machine takes care of all of it. And then at the other end of the spectrum, we might have basically just the human doing things. That’s sort of the baseline at the bottom.

[00:20:52] Martin Fischer, PhD: And then increasingly you could say, well, there could be support for the digital method. For routine types of work. There could be routine autonomy, there could be support for non-routine work. And then there could be, like, the system can handle the digital methods, the machines can handle anything. So if we map the examples that are shared onto this framework, we see that, for example, the…

[00:21:22] Martin Fischer, PhD: …data collection system for the crane offers plaques, basically data autonomously. And I think only in some, I would say very non-routine situations, is the human, does the human, need to look at what’s going on and add to the data. But most of the data is observed, collected pretty much autonomously in terms of the analyst role and the optimization examples.

[00:22:00] Martin Fischer, PhD: The optimization methods can handle. Sort of standard buildings, construction schedules, so extremely unique situations. Probably not yet, so that’s why I put it there. In terms of what? In terms of providing routine autonomy the virtual reality method basically provided support for decision making for a team of people.

[00:22:24] Martin Fischer, PhD: So that, in that sense, is a bit lower level in terms of what the digital twin is doing. And then in terms of the autonomous or action executor, the drone really handles all of the layout. The drone really handles all routine situations and actually provides support for non-routine situations.

[00:22:47] Martin Fischer, PhD: And I would, and the rebar tying robot handles routine situations very well, but none, or the complex type of rebar, designs, etcetera humans still have. So if we, while I channelize this, and the challenge you have and the opportunity you have in your companies is to decide of these form types of production tasks in terms of gathering, observing data, making sense of the data, using it to make predictions, making decisions, and then actually executing the action for a particular situation.

[00:23:23] Martin Fischer, PhD: How much does the machine do and how much does the human do? And so therefore, how do you combine the digital and human capabilities? And this particular example, you know, the line shows that an automated well data is collected with some support of humans, but the humans still have the major role in actually collecting the data.

[00:23:46] Martin Fischer, PhD: Then once the data is collected, routine situations can be analyzed automatically. And also decisions suggested automatically. But then the implementation, the action, is left to the humans and the human is handling all of the non-routine situations in terms of analysis and decision making and also quite a bit of the data collection.

[00:24:12] Martin Fischer, PhD: Then you could use that, as you say, well, this is what we have today. And then you could decide strategically, right? How do we want to change this? Do we want to create some level of autonomy in the execution? Do we want to be able to analyze non-ottin situations? Is it important that we gather data more automatically?

[00:24:30] Martin Fischer, PhD: This, in my mind, is how you can then connect this blend of digital and human capabilities with your operational and strategic. Because on the operation side, of course, we have to make sure that all of this is working. This planning of human digital capability is working really well and consistently from project to project.

[00:24:49] Martin Fischer, PhD: Otherwise, we know what our project teams will do with the tools. So I hope these examples and reflections have explained to you the opportunity we have with modern construction by really combining digitalization autonomy in this industrialization so that we can optimally blend, combine human and digital capabilities in our project production system.

[00:25:17] Martin Fischer, PhD: Because I would say we simply, we can, but we also have to. Thanks for listening.

[00:25:28] Gary Fischer, PE: Barbara. All right. Thank you, Martin. That was a lot to think about there. Before we move on and invite our other guests for the panel, I want to give you a 30-second plug on the partnership that we have with you around some education around these things.

[00:25:48] Gary Fischer, PE: So maybe you could just talk about that briefly.

[00:25:50] Martin Fischer, PhD: Yeah. So we have had the pleasure and K MO is a graduate of that program. With PPII and SPS for over a decade to deliver the virtual design and construction program created and delivered globally to thousands of people on every continent in every role.

[00:26:14] Martin Fischer, PhD: And really what’s needed now to move us further on, and that’s still a very relevant program, of course, but what’s needed now, which I hopefully was able to show a little bit, is a much stronger conceptual underpinning of how we produce work, how we plan it, how we manage it, and that’s where really I’m thankful for PPI to have made me aware of operation science.

[00:26:53] Martin Fischer, PhD: Until a few years ago, you could have said, well, operation science, yeah, Martin, that’s very nice. All Gary, you know, or, or Verto, great theory. I have a project to build here. So, you know, but now as you could see, we can get the data, we can analyze the data, we can bring some autonomy to our projects. So now you have a choice of just continuing with the…

[00:27:18] Martin Fischer, PhD: I would say pretty flimsy, conceptual underpinning of construction management or put construction management on a really sound footing, a conceptual footing. And that’s what we want to do together. Yeah, teach very well. We hope that if this sounds interesting to you get in touch with Gary.

[00:27:41] Martin Fischer, PhD: Me. Yeah. And we’ll work with you.

[00:27:44] Gary Fischer, PE: Very good.

Read More

Speakers

Read Biography

Martin Fischer, PhD

Stanford University

Martin Fischer, PhD

Stanford University

Martin Fischer is a Professor of Civil and Environmental Engineering and (by Courtesy) Computer Science at Stanford University. He is also the Director of the Center for Integrated Facility Engineering, a Senior Fellow of the Precourt Institute for Energy, and the Coordinator of the Building Energy Efficiency Research at the Precourt Energy Efficiency Center.

His research focuses on modeling, predicting, measuring, and improving the life-cycle performance of the built environment. He is known globally for his work and leadership in developing virtual 4D modeling methods to improve project planning, enhance facility performance, increase the productivity of project teams, and further the sustainability of the built environment. His award winning research results have been used by many small and large industrial and government organizations around the world. He has published over 100 refereed journal articles and book chapters and given over 50 keynote lectures on his research.

He has lived, worked, consulted, and taught in Europe, South America, North America, the Middle East, Asia, and Africa. His consulting work includes building owners, architecture and engineering firms, construction firms, government and research organizations, and software companies.

He holds a Diplôme d’Ingénieur in Civil Engineering from the Swiss Federal Institute of Technology in Lausanne, a MS in Industrial Engineering – Engineering Management and a Ph.D. in Civil Engineering – Construction Engineering from Stanford. He received the CAREER award from the National Science Foundation in 1996, was named a top 25 Newsmaker by Engineering News Record in 1996, won the best paper award from the ASCE Journal on Computing in Civil Engineering in 2002, and was elected as a Foreign Member of the Royal Swedish Academy of Engineering Sciences in 2012.