Atul Khanzode–Technology and Innovation Leader at DPR Construction–discusses digital twins, designing for efficiency, and deploying technology.
Gridium: Hello everyone, and welcome to this conversation with Atul Khanzode, Technology and Innovation Leader at DPR Construction. Atul has been on the DPR team for 22 years and also serves on the board of its venture capital arm. He has his Ph.D in Construction Engineering and Management from Stanford University with a thesis on the virtual design of Mechanical, Electrical, and Plumbing systems. This in addition to other engineering degrees.
My name is Millen, and I’m with Gridium. Buildings use our software to fine-tune operations.
Atul and I will be discussing some of the latest popular topics in construction technology. And there has been quite a bit of activity in this industry indicated most recently by a $131 billion investment level in the construction venture ecosystem in 2018—an all-time record high according to the National Venture Capital Association.
Atul, you and I met at a building technology meetup here in San Francisco, and given your extensive experience and expertise in our domain, I am just thrilled to have you on the podcast today.
Atul: Thank you Millen. I’m excited to be here.
Gridium: You’ve been a leader in the industry for two decades. What has changed the most over that span?
Atul: I think there are a few things that I can point out that have changed over that span. One is, I think buildings have become more complex. I think the way the construction industry or the wider AEC industry procures services of everyone from designers to general contractors has evolved. I’ve seen more collaborative approaches to procurement of those services and there’s obviously a lot more use of technology that is allowing us to integrate better, collaborate better with the supply chain. So, those three things, I would say, have evolved over the last two decades.
Gridium: Yes, and indeed my family designs and builds homes and small commercial buildings in New Mexico. And I’ve grown up on construction sites and have spent quite a bit of time with a shovel (Laughs), so I’ve seen some of that myself.
Navitas Capital, an investor of ours, points to two specific trends in their latest construction and technology white paper: (1) automation & off-site work, and (2) digitization of workflows and data-driven decisions. Atul, what do you see as the role of off-site work, robotics, and automation?
Atul: I think as technology has allowed us to basically coordinate the various systems better, the opportunity for off-site work has increased because of the confidence that you have in installing those systems that are built off-site.
So, I do feel like I think that is only going to go up. One of the ideas that I feel really excited about is the transition and going directly from BIM through robotics into fabrication and then building some of those multi-trade assemblies offsite, and then bringing in and essentially assembling these off-site multi-trades fabricated assemblies—bringing them on and doing them on site or installing them on site or assembling them on site. So, I think outside work is already a component of our projects—many of the mechanical, electrical and plumbing subcontractors have done that for their own trades, and we are seeing more and more of it with either sort of this multi-trade approach or sort of this volumetric-through-fabrication approach.
And I think that that’s an exciting development for our industry.
Gridium: Interesting. There is also a lot of thought going into construction industry productivity—is the digitization of workflows and data analysis making a dent in this trend?
Atul: I think it clearly is; I think productivity in this industry is how efficient can you be for your on-site activities.
And when you talk about digitization, the use of building information models for coordination, making information location-aware… I think having real-time feedback through laser-scanning and other approaches on how you’ve done installation; I think all of these are helping the industry improve productivity by making the workflow on the job site—which tends to be really sequential—go a lot smoother.
I think there is a lot that digitization through BIM and other tools have already done to improve the workflows, especially for the workers that are doing the installation work in the field.
Gridium: Does DPR think about any potential distinction between productivity and efficiency? For example, we could imagine a system optimized for throughput but, nevertheless, might have quite a ton of waste.
Atul: We do think about this quite a bit and one example that comes to my mind is: we’ve done a lot of in our recent larger healthcare projects where we have brought in a team of designers and general contractors, which is us—plus the subcontractors—all together in a big room to essentially design something and coordinate something to a level of detail to where the construction will go really smoothly because we’ve coordinated everything.
Now, that has lead to some amount of inefficiency for, let’s say, a smaller player who could be needed only to partially answer during the coordination process any questions that come up. So, for example, like a medical planner or others who have a smaller role to play, but when they do have a role to play, it’s a pretty important role. So, I do feel like we need to respect that and, in some instances, even pay them for their services and their time so that we can optimize the overall system and improve the overall productivity; although, if you look at the sub for a particular trade, for them, it might be slightly inefficient, but it’s better for the overall productivity.
Gridium: I see. Yeah, that distinction is interesting.
Atul, you also ran the virtual building group for four years. Can you describe the difference between a BIM model and the new term “digital twin”? And do you think this is a meaningful difference?
Atul: Yeah, so to me, a Building Information Model is an extension of a 3D visualization in a sort of object-oriented properties for all the objects that you can create and do further analysis with, like quantities or cost or schedule or other things.
The digital twin should represent a little bit more, so I look at a digital twin as kind of a living, breathing operating facility or a digital representation of that. And mostly, what I see is a digital twin. When I hear that more than likely people are referring to the product model with the properties associated with that product model—like the objects in the model—and I do not see the organization or the process represented in it.
And many times, I think that is a critical need when you look at it from a building operations perspective. Like, what is the fitness of the facility to the operations or how do you maintain it? What is the process you go through to do that. And to me, a digital twin would be richer if the process and organization views are represented along with the product view in the digital twin conception.
Gridium: Well, in speaking of that, what are the challenges—or is that it—to a single end-to-end data stream from design all the way through construction to O&M? And what might be the value of that?
Atul: So, I do see a lot of value in utilizing data from design and construction through the O&M process.
I feel like the challenge that we have is the fragmentation and, through the life cycle, the contribution of each of the players in making that useful for the owner. And how do you really manage that? So, an example of that would be: a design model many times doesn’t have the information on how something is actually constructed, but a constructed model doesn’t have information on things that a designer might care about; for example, the circuit information or things like that.
You know, they might have all the information to fabricate and install something, but might not have the system-level information. And all of these need to be brought together to make the models relevant from an operations and maintenance perspective. And to me, I think you need to start with the end in mind so that you can basically harvest this information through the life cycle as you are going through the design and construction process.
Gridium: Given your experience, can you speak briefly about what you’ve seen as the most important elements to owners versus designers, as compared to general contractors, as compared to subcontractors?
Atul: You’re talking about just the building information models? Or in general what their needs are?
Gridium: Good question. Let’s say both, but let’s start with the models.
Atul: What I see from an owner’s perspective is that they want to be sure that between the designers—let’s say the design build or design assist build or design assist contractors and the general contractors, that they have bought all the scope—that there’s no scope that’s missing that they will have to pay for.
I think from a designer’s perspective, is it detailed enough to meet the requirements of what they need to do, which is to produce the formal documentation and to have the clarity of what is being requested or specified that needs to be built. And then from a general contractor and subcontractors’ perspective, is there a clarity so that we can go out and I think what we have done in the building information modelling of what will be done in the construction process.
And go and fabricate it so that our construction process is a lot more predictable and we are not having to do the rework and we can track productivity and we can come within budget and time. So, I think that’s what I feel the various players involved in this process really care about.
Gridium: I wonder if the digital twin might be that data layer that brings together and unites the stakeholders’ differing priority lists? Do you think that’s possible?
Atul: Yes, I do think that’s possible and as I think back about the life cycle, at different times in the project life cycle, different things within the digital twin become more or less important.
So, I think doing the construction process for example—a lot of what we’re looking for is, how do we plan construction activities and how do we essentially be able to fabricate from the models? Things like that.
I think when it goes over to the owner, they’re really more interested in understanding the system-level view so they can do things like: in case of an emergency or a leak or something, that they can isolate a subset of a system or things like that. So, that then becomes more relevant so I think absolutely, yeah. I think if a digital twin represents all that information through the life cycle, then absolutely it’s a good concept to have.
Gridium: How focused are owners on total cost of ownership, or designs with efficiency as an objective function, or plans that account for preventive maintenance?
Atul: I see two distinct levels of our engagement with owners, especially when we are dealing with institutional owners. They’re really, really focused on the efficiency of their building operations, how they’re going to have their preventative maintenance program, how their assets work together with other assets they have—they’re very focused on it and very interested in this whole concept of digital twin or facility operations models, or other things.
And then there are other models who maybe are building for the first time who do not have a huge investment in facility management organization that are probably less sophisticated, in that regard—they’re interested, but maybe less sophisticated in that regard. But I do believe that everyone that we work with is very interested in obviously getting a quality building for what they pay for, and then to have the ability to be able to manage and operate that building most efficiently.
Gridium: Sometimes General Contractor Research & Development is described as comparatively quite low to other industries. Do you think this is a perception, or is it real?
Atul: If you look at the absolute numbers and compare ourselves with other industries, it is real. And it probably has to do with how much of a low margin industry we are. And also, to be real, I think a lot of the investment that is necessary in technology and other things to be brought to our industry, I’m not sure that only a single firm within our AEC ecosystem can pull it off. So I think, it does require a different level or thinking about R&D and innovation in our industry.
Gridium: How does DPR manage innovation and entrepreneurialism inside the firm?
Atul: We have a few things that we have invested in. We have an internal innovation group that we constituted about 10 years ago.
The main objective of that group is to fund pilots on either process improvement innovations—various things—that sort of sits outside of the project budget. So, we fund this from our corporate innovation budget.
And one of the other approaches that we have also taken is to really partner with entrepreneurs that are trying to push the industry, that have invested on their own in various tools or technologies or processes. So, we want to get them engaged through funding these innovation pilots on our projects. And when we look at that, we basically have a criteria on how this helps the owners—how does this help become a better technical builder?
There’s a few criteria that we have that we evaluate this around. And then, on the entrepreneurial side, what we do is we have considered a venture fund and when some of these ideas have bigger legs or need a lot more funding, then we have funded them.
We have funded a company that’s focused on essentially building operations that’s a software service company, and that’s basically focused on providing a platform for building owners to manage and operate their buildings. We also funded company that is focused on digital fabrication and they’re operating as independent entities in the software fabrication space.
Gridium: Can you tell us about your experience deploying new technology on the jobsite? How does your team monitor engagement and measure results?
Atul: We usually tend to engage with our project team, so we have a group within DPR called “Technology Integration Managers.”
So, at the beginning of any project, we sit down with the project leadership team and really go through—without first talking about technology—all the various processes and the way they’re going to manage that. And then bring them basically the solutions that include both process and tools or technologies that are available to them in order to sort of deliver the process to the owner on that project.
And then that group works together with the project team to bring the relevant technology on the site, and that might include not just sort of setting it up for them, but also training them and helping them through the process.
Many times, through the innovation budgets we would also fund something new if they want a try out something new on the project that either they might have been exposed to or they can also look at our innovation portfolio to learn about something that we might be doing that they may want to try out. I think it’s a much more sort of collaborate approach within DPR.
In terms of monitoring engagement and measuring results, some of the things that we look for—especially for the pilots—is doing check ins on a periodic basis through the innovation leaders of the business unit or the region to understand whether the promise of something that we wanted to do: is that coming true or creating more work or not?
So, I think we do multiple sort of checkins to really understand how this is going. And I think we can get better at that. Sometimes we don’t kill stuff that’s not working soon enough, but that’s kind of how we do it.
Gridium: Are there any examples where a new technology simply did not work, and what are the positive opposite case where something really delivered value?
Atul: There’s a bunch of examples that I can think of where something didn’t work–we had a health system that was based on this lean idea of having a conversation which is kind of mixing the message and having somebody really understand this network of commitments kind of a system. And we tried mightily to be integrated with our email system and our telephones and we tried mightily some more but it just never worked. So we had to abandon that and go back to the old way of just bringing people together in a big room and sort of work through that.
And then there’s others that sort of really worked. I think having technology on iPads, various things like Bluebeam, PlanGrid, stuff like that that our field really likes for example that has worked. We don’t have to push it that much. I think people see the value in it and they have information at their hands, like out in the field—you know, things like the computer on wheels where we have it in a jobsite box out on the floor where construction is going on. I think those things really work.
Gridium: Is there anything that you’re really excited about this year as you look ahead?
Atul: The one thing that I feel excited about this year is I think on a couple of our institutional projects, our large healthcare clients, the owner taking interest in the O&M system that we have deployed on their project, and the value that they’re seeing from it.
It gives me a lot of satisfaction that we’ve not just built and delivered a great construction project for them, but we’re also continuing to be engaged with them in really delivering the service that they as a world class kind of healthcare facility provide to patients and that we are playing a small role in helping them manage their facility. I think that’s really been pretty exciting for me.
Gridium: Cool. Well, this has been great, Atul. I really appreciate you taking the time today. Your thoughts here have been quite interesting. Thank you!
Atul: Thank you very much, Millen. This was really awesome to talk with you.