Architecture, And: Robotics

A This episode of Practice Disrupted is supported by Monograph, the cloud based practice operation solution built for architects by architects. I'm Evelyn Lee. And I'm Janine Chastain. We're collaborating on curated conversations to explore how the industry is changing. Together, we'll find ways to create new solutions to current challenges while elevating the value of architects. Welcome to Practice Disrupted. Hi, listeners. Hi, Janine.

A Hi, Evelyn. Hi, listeners. I'm usually the one who's playing in the text space, Janine. So because you're our primary connection to the guests, what made you so interested in bringing Brian on?

B I wanted to bring Brian on the show because I wanted to understand his career. He found his way into a full time job at Boston Dynamics, and I'm curious. How does a degree in architecture lead you to a career at Boston Dynamics? Boston Dynamics is a world leader in mobile robots, tackling some of the toughest robotic challenges. They combine the principles of dynamic control and balance with sophisticated mechanical designs, cutting edge electronics, and next generation software for high performance robots equipped with perception, navigation, and intelligence. So before you listen to this episode, I think it's a good idea that if you scroll down to the show notes, you're gonna find a YouTube video that takes you to Boston Dynamics and a video called do you love me? Take a minute to watch this video because I think it will help you really fully appreciate the conversation that we're about to have and why I'm so excited about it. In addition to that, there's a 2nd video listed called meet spot. This is the specific robot that Brian who's going to be speaking during this interview is supporting in his current role. Spot has the ability to walk upstairs.

B It can avoid obstacles, and it can navigate rough terrain like loose gravel. So some construction sites are deploying Spot on job sites to automate inspection and capture data in live time in the field, And Brian's gonna explain this and more in the interview.

A Before we jump in, a little bit more on Brian. Brian Wringley is a construction technologist at Boston Dynamics, where he promotes new value add autonomous capabilities for construction project delivery and works to expand the construction application ecosystem with the Spot SDK. Prior to Boston Dynamics, he was a senior construction automation researcher at WeWork, where he managed the construction robotics research program, and contributed to initiatives and design automation, unitized prefabrication, and construction site progress monitoring. He has also taught architecture courses for many years, most recently at the Pratt Graduate School of Architecture and Urban Design, where he led seminars in computational fabrication robotic automation for industrialized construction. Let's cut to the interview.

B Welcome to the podcast, Brian. Thanks so much for doing this. To get started, why don't you tell us a little bit about yourself, where you work, and maybe just a little bit about your background?

C Yeah. Thank you so much for having me. I'm currently a construction technology manager at Boston Dynamics where I work with SpotOn customer environments like construction sites. I was educated as an architect. So prior to that, I had worked on a design technology team at Woods Baggett and had also worked as a construction automation researcher at WeWork, where I was actually an early customer of Spot.

B Yeah. You have a really unique blended background, which I think we're gonna get into a little bit further on in that interview, but, I wanna first start some of the people in our audience will know what Boston Dynamics is, and for some, this will be their 1st introduction to the company. So can you tell us a little bit more about, Boston Dynamics and what you do?

C Yeah. It's the world's most advanced robotics company, and I think most people know about us through our YouTube videos. So check those out. The history of the company is interesting. It started as something called the Legge Lab, first at Carnegie Mellon in the early eighties, and then on to MIT before it was spun out as a private company in 1992 by our now chairman, Mark Raebert. We went through a series of exercises bringing some of the very first legged robots out of the lab and into the real world in the early aughts with robots like BigDog and l s three. The company was acquired by Google a little bit thereafter where the spot project was started to think about how to bring quadrupeds into more human purposed environments and to think about something that could be mass manufactured and useful for a variety of industries. The company was then sold to SoftBank, and then most recently, SoftBank sold a majority, approximately 80% share of that to Hyundai.

C And that's where we stand today.

B You shared this story when I listened to the interview on Troxell about the first time that you saw a spot on a construction site. Can you tell us about that memory?

C Yeah. Yeah. Because it was kind of in the context of the work I was doing at at WeWork. We were looking for ways to automate data capture on our job sites. We had acquired a couple general contractors in different regions of the world, and, you know, surprise, they didn't have very robust data capture programs. So we're looking for a way to do that, and we were pretty limited by using our supers and assistant supers to go around with handhelds to be able to do that at the frequency and repeatability we really wanted. So I'd spent, like, nearly a year looking at wheeled and tracked vehicles and the use of drones and doors and things like that and just kinda hitting my head against the wall because none of it really worked in a dynamic environment like a construction site. And in my original report that I had given to my team, I had noted the existence of lagged robots like Spot and, some of the stuff some of our, you know, friends at, like, agility robotics were doing.

C And, you know, there were a couple of those things I was just like, this is this is so exotic. Like, we would never there would never be justification or use for this on a construction site. So jump cut to a year later when I'm, like, pulling my hair out, and then this video comes out on YouTube with the alpha spots, the black ones walking around Japanese construction sites, which are very clean. You'll notice in the video, you'll be like, that's that doesn't seem like a real construction site that seems staged, but it's just the Japanese construction sites are so immaculately run. And it had a an early version of the arm with a camera and the gripper. It had a 3 60 panel camera and pan, tilt, zoom camera on the back. I was going around and and taking photos. So I immediately, you know, asked everyone I could to find a connection to Boston Dynamics.

C And, like, a week later, we went up on the train and and had our 1st conversation.

B Did you end up adopting them into your projects, or or were you able to convince your team to use them?

C Yeah. So we were. We ran a pilot on a project in downtown Manhattan, and we were able to prove out quite a few things. I mean, 1st and foremost, we wanted to prove out mobility on the site. Can this thing actually move through a typical floor plate? Can it move up and down stairs? We had a lot of contiguous sets of floors that we would lease and then build out into WeWork spaces, so that was an important aspect. And that worked. And then we said, can we do it autonomously? Can we record this path and then replay it and allow the robot to exhibit things like obstacle avoidance, you know, not run into me, be okay if it steps into an empty bucket, things like that. And that really worked out pretty well.

C And then we said, okay. We have this construction progress management application that we had also acquired called, FieldLens. So can we use the Spot SDK to connect to FieldLens APIs to upload the data using our job site Wi Fi so that images collected on job site would be, you know, transferred in near real time to those systems in the cloud so that other stakeholders could access that information. And that actually worked too. So we were able to tick off all of those boxes, like, pretty much immediately, and that's when the light bulb went off. Like, oh, this might be the future of of the way we manage construction sites. And for me, personally, in my career, this might be the way I could have the greatest impact on the industry.

B There were a few reasons I wanted to bring you on the show. First of all, I'm a huge fan of Boston Dynamics as are you. But, also, you wrote this article that, came out in architect called master builder, how robots can really transform the role of architects. And I I would love if you could give us, like, the high level premise of your position in that article and, obviously, it's probably informed by a lot of the work you're doing.

C Yeah. Absolutely. The position overall is the way that we educate young architects to think about robotics and automation. It's often in the context of the mythos of the master builder. Right? This idea that a preprofessional architect centuries ago was somebody who did the design, but also did the actual building of a job. So I wanted to say that there were a couple reasons that was both problematic and just didn't really work well with the way business models are done today, the way contracts are written, the way risk is managed, who's responsible for means and methods, you know, reality check and said, that's okay. We don't need to be directly controlling robots to, you know, 3 d print concrete on job sites with our grasshopper scripts. That's not that's not maybe the place we could be most effective.

C That's a great way to teach those concepts in terms of design for manufacturing and machine readable deliverables, all sorts of things like that. But ultimately, we need an infrastructure by which instead of obsessing over putting all of the design intelligence into a model and pushing that, you know, down the throat of the delivery process as much as possible. We need to develop a feedback loop. You know? The issue is that we don't know what's going on on-site as designers, and our work suffers because of that, especially targeting that construction administration phase. Beyond the construction administration phase, I started to hint at the fact that right now, there are a lot of owners interested in having more sophisticated, digital assets to manage their projects, the digital twin. Right? Whatever that means. But I think the digital twin is is ultimately an owner asset. But what we can do for owners at handoff is prepare something that is an as built that is up to date that has three-dimensional information and information, so BIM, plus historical data about the progression of the job site and critical things like, record of what was in a wall before it was a before it was closed up, a record of the rebar layout before the slab was poured in case you need to make a penetration in the future, all sorts of stuff like that.

C If you can give that record of how the building was constructed combined with an accurate BIM, you know, that's really been calibrated to the reality capture over course of the project. That's the foundation for the owner to take that and develop that into a digital twin. But the truth is that that piece was something I wrote out of guilt because I had spent years years educating architects to try to build these models that the universe down to the model and to use that to directly drive and roboticize all construction activities. And then I was humbled at WeWork because it was vertically integrated. We were truly able to go end to end. And I saw all the pieces that I had missed, and those pieces were people. And those people were highly skilled trades on job sites. And I realized we don't we don't need to have everything figured out.

C We need to encapsulate the knowledge of these downstream trade teams and just establish a feedback loop. This does not have to be a linear one way process.

B I think what's really fascinating is just your thought process around the business model of architecture, but more specifically, like, diving into the construction side. I think this is the 1st conversation that we've had where we're really talking about that that phase of the process and, like, lessons learned about where to make improvements and how to change the way that we're thinking about construction as a whole. So maybe we should step in further to this idea of of robotics on the construction site. What can you tell us about where you think the future of that is heading?

C I see a couple distinct things happening, in terms of the types of robots we're seeing on the job site. Something like Spot is an agile mobile robot that's capable of taking sensors autonomously through human purposed environments. There are a lot of advantages to that, but the focus is very much on sensing. And the idea that reality capture is not something you do on a job site, you know, once, maybe twice for for due diligence or kind of a final as built because it's so costly. And labor intensive. It's something that you now have the technology to do continuously to establish that feedback loop and to give everybody up to date information. There are so many reasons for that to automate quantification of work in place, which can be used downstream to automate payments to identify mistakes earlier and minimize rework, RFIs, change orders, you know, you name it. There's the the fundamental problem of construction is that every construction is over time, over budget.

C So anything you can do to chip away at that is helpful. So I think that's that's the fundamental thing that Spot is offering. It's also I think there's a conversation too about the idea of location services on a job site. There are lots of ways to do that. You could have beacon systems. You can use the phones in your pocket in certain ways. But how do we locate what's going on and track material movement and activities through the job site. So the more we have to think about the real world in terms of time and location and then attempt to kind of correlate that with the design model, I think the better off will be.

C That was a a huge lesson at WeWork was they had a really skilled reality capture team, and I learned what it really takes to adjust a design model to the reality of the site and really left that experiencing no longer trusting architects' models and saying that is a that is a design intent artifact, but has no bearing on the reality of the situation to the point, you know, we have a lot of conversations and, you know, customers are like, I wanna drive the robot with with Revit. And I say, would do you? Because I don't know if I don't know if that information is accurate enough to control these downstream processes. So, you know, back to that article, if you do actually wanna think about automating job site tasks where you really have to be in the right place and have a sense of exactly where that machine is and the object it's trying to place, you need to really establish that handshake between the design intent models and then what actually gets built. So that's what gets into that reality capture feedback loop. Now there are a number of other robots that are already just jumping ahead and thinking about doing one specific task really well, and a lot of them came out of stealth this year. Dusty Robotics is doing layout parking, and then you have Canvas doing drywall finishing, and then you have the Hilti robot doing ceiling drilling for for anchoring in MEP systems and hangars and things like that. So you have this emerging group of task specific robots, and I think that's good. I think one thing you should ask when you're implementing task specific robots and robots that start to do the work of the trades or augment the work of the trades rather is to say, there's a huge push toward industrialized construction and off-site construction methods.

C Architects and designers are recognizing this, and they're thinking about the ways that they can design differently to build elements of their buildings and factories and manage the logistics then of moving that to a construction site in the field. So if you do the thought experiment of everything I design gets built off-site that's possible to build off-site, what is left to do by robots? Well, you get data capture and progress management spot. You get layout, which is dusty, and to a certain extent, the Hilti robot. I mean, a lot of what it's doing with placing anchors is really a layout exercise. You get placement of objects. I don't think there's a, you know, a a robot that's really placing things right now. There are some efforts around automations of of both cranes and mini cranes and things like that. And then you get finishing, right, with what Canvas is doing.

C So I actually think the companies that are out there doing the task specific robotics chose the right things with respect to the parallel push for industrialized off-site construction methods. And ultimately then, all of those things will be working together. You'll want all of the robots working off of the same location services. You'll want them to perhaps be able to interact with one another as heterogeneous fleets to achieve tasks that are more complex than what any individual robot could achieve itself.

A I wanna take a little step back because you talked about kind of just your evolution of the companies you worked for to get here. But I think your path in itself is intrinsically unique, and you came out of what is, you know, seemingly an m r c program. So how did you find yourself in this space, and what was your journey specifically to robotics?

C Yeah. I mean, that's a great question. I I knew, like, from a young age, like, in high school, early in high school, I wanted to be an architect. And I had a fascination with with modeling technologies, like drafting and and three d modeling. So there was a little bit of a technology combined there. But as I moved through architecture school, I really just wanted to be a designer, and I wanted to go out and practice architecture. So I think the fundamental thing that happened to me was the recession. I graduated in 2009.

C And after the 50th no to work somewhere, I said, well, I'm gonna have to do something else. I can't practice architecture right now. So I had done a fellowship at the University of Cincinnati in what was called the rapid prototyping center, which was like a digital fabrication lab for primarily for their industrial and transportation design programs. And I said, what if you let me keep working here, because I have nothing else I can do? And what if I served because of my background in architecture, what if I served as a liaison to bring in the architecture, interior design, and urban design programs into this facility and really because I really think they could they could make amazing use of this. I had done some work in grad school where I had tinkered in there, but I noticed that the architecture school as a whole was still pretty hands off with some of that stuff. There was a strong, at the University of Cincinnati, I mean, there was a strong culture of making, but they hadn't they hadn't really been attuned to the digital fabrication stuff just yet. So that pushed me into that role and then teaching in parallel. And the story I always tell is that it's happened is that the grad students at the time had a petition because they were angry that no one was teaching Grasshopper.

C I mean, Grasshopper was probably, like, 1 or 2 years old at this point, but they they were on the ball. Right? And they knew that that was the future, and they wanted that skill. So the director of the graduate program, you know, picked me as the young computer person on faculty and had me teach that. So then I had to teach myself Grasshopper, which was actually great. That was a good, like, forcing function. Started teaching that. I really wanted to go to New York eventually, you know, for the architecture scene there, primarily. Finally made my way there as part of NSF grant, in a CUNY system.

C And it was a cool opportunity too because instead of working for an already established digital fabrication lab, I could build 1 from scratch. But we were also designing a curriculum that was not subject to NCARB, which was awesome. So it's an unaccredited program, you know, at the time. I I think people have very different opinions about this, and it's a big subject of debate. But City College in the CUNY system had the, like, you know, that was the NCARB certified accredited MArch pathway. And City Tech in Dumbo in Brooklyn was really there to say, well, how do we reinvent ourselves? We're serving a huge population primarily from the boroughs, and we wanna give them these nontraditional career pathways that have to do with architecture, but don't have to be practicing architecture itself. It it could be a way to get to an, you know, at city college or elsewhere, or it could just be a way to start working in an office where you're doing, you know, sustainable engineering, or there are so many really amazing architectural fabrication facilities in Brooklyn. So I was really captivated by that mission and the flexibility.

C We made a really cool program where you could get capstones in computational fabrication, BIM, and or, performance driven design. And we had seminars where students in each of those tracks, like, were basically pretending to be different, you know, stakeholders working together on the same project, sharing, you know, models on the cloud and and running advanced analysis. It was really an incredible program. I I was like, I wanna be in this program as I was teaching it. And then because of that, we were also taking on advisement from practitioners in the city. It was very much oriented, not necessarily strictly toward our potential practice, but toward all these professional opportunities. So we had people from the city always coming in, giving lectures, informing the curriculum. One of them was Shane Berger, who was running the design technology team, was Baggett, and then there was an opening on his team.

C They had just done the SAHMRI building in Adelaide, which, like, really blew my mind in terms of what you could do with, design computation to make a really innovative and functional space. So I got that job, and that was like my dream job, right, to finally get into the industry, but to be practicing in this kind of contemporary way. And then, you know, 3 years later, you know, still very happy. I mean, that was an amazing experience. I'm still in touch with Shane, and I actually think Woods Baggett is, you know, still one of the most cutting edge practices in in the world. But what I realized fundamentally was that across the industry, there was only so much you could do when you're building design computation models. So we were working on a number of high profile projects, including 76 11th Avenue, the twisty facade for a new project still under construction on the west side of Manhattan. And we're building these really sophisticated models to automate the generation of the cladding as well as, you know, a lot of the interior elements that were dependent on that cladding and this, that, and the other.

C But then realizing that a lot of that stuff just wasn't surviving the scope, and that we were just cutting the model off at the head, issuing drawings, and we're unable to leverage that model for all the QAQC we could have during construction administration. So that really troubled me, because I felt like I was I was making my life easier for sure. But I I wasn't sure if I was making the project any more efficient overall. So that's when everyone I had ever known had been poached by WeWork and was like, alright. This seems like a you know, it's worth $40,000,000,000. That can't that can't go wrong. And everyone's there, and it's vertically integrated. And I remember somebody who's working there at the time was like, yeah.

C We're looking into, like, layout robots. And they and I was like, what? But it's like I was like, you mean there are robots other than just, like, robot arms and stuff like that. So, like, that really piqued my interest. Joined that team. It was incredible. It was such a cool experience to be able to work on design automation software in the morning and then in the afternoon to be able to, like, go to the factory that was spitting that out from specialized CNC machines, and then in the evening to be on a construction site, installing that, laying the site out, and testing robots for data capture. It was incredible. And you you really saw, like, oh, you can do a lot of this, but it also showed me some of the fundamental roadblocks that had to do with things like, how do I integrate with ERP systems in my building factory that's driven by my model? How do I make sure that the reality capture bottleneck is moving fast enough? How do I incorporate the trades? Because you're never, like, fully vertically integrated nor do you really wanna be because you, again, you don't have that expertise.

C That's that's hubris. Right? So that really pointed out all those things. And with the use of the robot, I thought, you know, it's again, it's not about this, like, linear extension of a designer's control into the delivery as much as possible. There are aspects of that, you know, for the good of the project, but it's really about developing some kind of feedback loop with site data. So it was not planned. It seems like that was a logical thread from one job to the other. It was very much about opportunities popping up and and really making sense in terms of how we could use these technologies that I love and fascinate me to really drive a broader impact on building delivery.

B I kinda wanna get you to talk more about what you're doing now in the construction technology manager role. Can you tell us, like, what a day to day experience is like for you there?

C Yeah. It's getting up in the morning and being like, why do I have so many emails? And then not getting to do any work because I'm just answering emails all day now.

B That must be the manager part of your job.

C Manager part

A of your job.

C Yeah. It's so before the pandemic, it was largely especially when we were first, basically white glove delivering robots to the early adopter customers, it was awesome because you were just flying around the world delivering robots, training up teams, and learning from them. Right? I was hired at Boston Dynamics because of my background and expertise in construction, but we try to stay really humble and learn from the industries that we serve. And there's just nothing that beats being on the job site with people using the robot and seeing exactly what's happening. So I miss that. We've, you know, we've taken a break from that. We found ways to work remotely. So a lot of it is still checking in on a regular basis with some of the early adopters.

C I'm happy to say we now have too many construction customers for me to check-in with all of them regularly, which I think is good, a good problem to have. But I would say most of the time is spent talking with them and understanding their feedback. And then, of course, the product management, the pure product management role of that is incorporating and structuring that feedback in such a way that that can get picked up and prioritized correctly by our engineering pipeline. You know, now that we have this product out in the world, we are updating that spot software and supporting systems on a regular basis and making sure that those updates are responsive to what customers are asking for in terms of, like, both the behavior of the robot in challenging environments like construction sites as well as the autonomy layer on top, which, you know, is a huge part of the value proposition, which is I need this robot to be able to sense this environment and move through it unsupervised even if this environment is is changing every day or gets flooded or it's snowing in sideways or people are in the way or somebody left a scissor lift out. You can imagine all of the crazy things that, these robots go through because we go through them on these sites. And then there's there's also stuff like this. Right? You know, speaking to other people in the industry about the possibility here, making the narrative clear about how these robots are are valuable and how they interact with other emerging technologies in the space. And then lastly, it's also about working with ecosystem partners so that they can understand how to use the Spot SDK.

C So we have a set of APIs that, you know, customers who are trying to develop software applications that consume spot data or work with spot data, or maybe they build sensors and hardware like laser scanners or 3 60 cameras that they wanna be able to attach to the robot and have be controlled through the robot autonomously. So trying to make all of that really, really easy for developers to do, so they don't have to sink a lot of time and investment into that process, and then can create value for what would be our joint customers.

A Interesting. I also find it amusing that while Spot is probably, like, you know, is, like, taking away some of the, like, the people work of of doing things on the construction site from a measurement standpoint is probably attracting a lot more attention too. So you actually probably have more people on the job site just coming to check out what Spot is doing.

C Yeah. There's a huge dip in efficiency that 1st week on the site when everybody's just ogling the robot. But it's actually that's a that's a fun moment on the site too because people get to share those photos with their families and be like, look what I get to do at work today. It's it's really cool the, you know, the tools that are emerging for people on these job sites. But, yeah, I think that I think that Spot really benefit you know, people are always like, oh, the business model problems that I think really center around shared risk, in the way the stakeholders deliver these projects. But, like, there is a really mature technology ecosystem. I mean, all these stakeholders have really strong priors in, like, three d modeling and, location services and cat reality capture and data pipelines and interoperability. Like, you know, that stuff has all existed.

C And there are amazing handheld cameras, things like the Ricoh Theta and the Insta three sixty. There's amazing survey and scanning hardware from the likes of Trimble and and other people in that market. So I think we've really benefited from that. And, yes, from that then, from that understanding, like, yeah, we already see the value of using cameras and scanners, and we see we know how to use that data downstream. We just need a better way of capturing it more frequently, and it's that frequency of capture. Right? That's that's the fundamental thing that the robot is allowing. And that's what opens up these other doors, which is, you know, every company that ingests reality capture data also wants to be able to update BIM elements based on that data or use, computer vision models to recognize a sheet of drywall from a 3 60 photo and tell you that that piece of drywall has been placed in your floor plan. Right? So they're all these emerging AI technologies essentially that can automate progress quantification.

C And that that is really where I think a lot of big margins and and opportunities for value lie. But to get the level of data and the frequency of capture where you're getting information same day so you can actually act on it, that's really where where Spot comes in. If you couple all of those things together, I think it's really great. And then, yeah, and then you get, you know, software that's being built specifically to control the robot. You've got what Hollow Builder has done with their SpotWalk application. DroneDeploy has an integration, that was developed with Brasfield and Gori for their 3 60 walk through features. So you have drones and legged robots capturing data from the exterior and the interior respectively and bringing them into this integrated, you know, 3 d reconstructed environment, which is super cool. Trimble is building out a solution where the robot is controlled within Trimble FieldLink, which is the software that people in the field are already using to run scanners and connect to GNSS receivers and total stations.

C So now you're adding that robot control into the mix, which means you can do the capture, and then you can also be doing the registration of these point clouds all in the field. So all of a sudden, you can do all this stuff you would normally have had to offload to an office somehow, like giant files that would had to have been sent via laptop or very slowly over a terrible network connection to somebody in an office to do something with just to get it to the state where it could be used by some other stakeholder. Now you can kinda prepare all that stuff in the field at the moment of capture. We have the Spot doc now, which allows the robot not just to recharge, but to also connect to a gigabit Ethernet connection. So you can do these huge data dumps, basically, like, have spark go around for 90 minutes and scan as much as possible and then download while it recharges and then just keep doing that. And by the way, if you wanna do that 20 four seven, you know, be my guest. So there's it's an incredible opportunity. And I think specifically with the release of the doc, we're gonna see that really unlocked in 2021.

A Let's take a break from this conversation to talk about our sponsor of this episode, Monograph. We're proud to partner with Monograph because they are helping to transform the practice of architecture, one design studio at a time.

B Tired of using dated and clunky software to manage your firm, or do you feel frustrated wrangling all of your spreadsheets to get

A a clear view of where your project stands today? Monograph is here to help. Designed by architects for architects, Monograph allows you to track your time, your projects, and your budgets in real time. With our awesome MoneyGantt, you can immediately understand project performance across your entire firm portfolio.

B Need to adjust your projects week to week? Their new tool, resource, allows you to reallocate your team's time and track its impact on your remaining budget.

A Be proactive with Monograph.

B Can you share a case study that you think is a really great example of this technology in use?

C Yeah. I mean, one of the earliest ones I thought was the most interesting, and I refer to it quite a bit, which is the general contractor Swinerton came to us. And then they weren't like, I want a robot because it's cool even though, you know, the robot is cool, and that's okay. But they came to us and they were like, we have a business problem where we're not paying our subs fast enough. And we think not only can we rectify that problem, but we can actually find ways to pay them so fast that we can incentivize more competitive bids, we can perhaps charge for those services from owners. Like, there's there's a huge opportunity here. That's driving a conversation which says, how do I automate payment systems? Well, I need some trusted metric around, progress quantification or work in place quantification to drive that. And, you know, so 50% of the drywall is up, and I can pay my drywall up.

C In order to do that, I need some robust data capture program and walking around with a handheld or a fixed system or, you know, various other means and methods just aren't cutting it. We think that the robot could be of use here. And, specifically, what we want is a laser scanner on the back, and then we need a piece of software that can take that point cloud and do a handshake between the features of that point cloud and my BIM elements such that I can now add parameters to those elements that say, was this object placed or not? Was it placed in the right location? Does it intersect with anything else? And was it placed on time relative to my schedule? Because I'm a very good BIM modeler, So I definitely have all of that well structured in my model. Right? Some of these technologies and opportunities actually push us to have better craft in the way that that we design. So they were using Avere, which is an interesting startup. Lots of ex we workers there as well. That's how I kind of got tuned into them to do that handshake essentially. So then they were like, okay.

C Now we have the opportunity, continuous data capture, and we're going to be updating BIM, which is great because I think there's a huge set of general contractors that are really striving to work with their stakeholders to push BIM downstream through the delivery process to the point where you could actually have, you know, some semblance of a proto digital twin to hand off to the owner. And that said, you have lots of other GCs that don't do that for many reasons. You know, BIM didn't survive the contract, and they're going to use really intelligent tools like the ones that are found with our partners, Hollow Builder or with, you know, their competitors. You've got open space, structured site, like 10,000 other ones. Right? Those are probably, like, the big three, and they all have, like, these competing AI tools that are able to pull building semantics. Like, I recognize types of building elements relative to the trades, and I'm gonna, like, demarcate that on your two d drawing set. And that's what's useful for you. So you can kind of see there's a there's like a little bit of of a split there, like, where BIM survives the contract, there's really this onus on high quality three d data and laser scanning and point clouds.

C And where BIM doesn't survive, you've got these 2 d document sets that you're trying to augment, and the simplest way to do that is with 3 60 image data. And then there are, like, you know, 10 other flavors of that kind of in between to create a spectrum of of possibilities there.

B So a

A lot of, I mean, there's a lot of scale happening here, and like all the contractors you're mentioning, you know, are working on projects of a certain scale. Where does robotics fit? If I'm a residential architect, and I'm, you know, and I'm thinking thinking, you know, where where can I drive, like, what does this mean for me? Does it mean anything for them?

C Yeah. I think that's a good question. I mean, I had some really interesting conversations with people building housing, for example. There's a use case that makes a lot of sense to me, which is that if you have a standard home design and that is the map from which the robot operates, then you can deploy that robot in lots of different houses, and it can recognize them because that's a repeatable design. If you put a robot in, you know, if you're building a large housing complex, I think the the challenge there is that, yeah, it could do a lot of the same things with construction monitoring that it does on, like, larger commercial and civil sites. You just got, like, a lot more doors in the way, I think. It's, like, one fundamental difference, and you're probably moving from the inside to the outside a lot. So it's definitely not the place where I'd say the robot has the most value today, but it's an intriguing conversation, and people keep asking about it.

C So there's definitely interest and utility there. The robot is, of course, capable of opening doors with the arm. The question is just, you know, what does that actually look like if you're operationalizing it? A a story about having a doc that lives on-site, a large commercial site and having a robot that's going around 1 or several floors on routine missions, opening and closing all the doors and other robots following through with, like, scanning and monitoring technologies. I think that one's a little bit a little bit trickier just because of the nature of that, like, technology. Right? So I think that's I think that's a challenge, but the fact that people keep asking about it signals to me that that there will be some there will be some way to add value there.

B I think, like, a lot of these innovative technologies, like, they they start with the large projects because that's where the budget can support it. And maybe in a decade, I don't know, maybe there'll be a way to scale it to other types of smaller projects, but I'm sure they'll discover a use.

C Yeah. I think, you know, hybrid indoor outdoor localization, something that can switch from, like, you know, some flavor of SLAM indoors to some flavor of GNSS outdoors coupled with the advancements we're seeing in remote manipulation with the arm. It it definitely seems like something that that could become very feasible in the near future. But that's just Boston Dynamics. Right? There are all sorts of players out there thinking about these things. So we could see something we didn't anticipate from a drone. Right? Some new feature that makes that safe and operational in that particular interior environment. We could see a proliferation of IoT sensors that make sense to include during construction that can actually survive then to becoming, like, smart home systems after completion.

C It's it's hard to say, but I suspect that you'll see something happen there.

B And speaking of Boston Dynamics, I feel like I'm gonna if I don't ask this question, I'll be sad later. So I I'm really curious, like, what is it like? What's the culture there like? And maybe this is really a pre COVID question, but I'm curious, like, what the culture of a innovative robotics company feels like when you work there.

C Yeah. I mean, it's, yeah, it's really amazing. I remember when I first went there when I was at WeWork and we were visiting for the meeting and they took us on the tour, and it was just like the weird quirks of, like, signs being posted, like, hey, a robot might walk through here, you know, watch out. And then, like, a spot, like, you know, trots by as part of some test mission. And then walking through the lab and just seeing, like, dozens of robots, like, splayed out in, like, the most grotesque positions just, like, on the ground, because they're in the middle of, like, some test. So that really, like, blew my mind. But no. It's it's really it's really amazing to work there.

C There's, I think, a really great culture of, you know, free exchange of ideas and collaboration. And to take like, I think what I like the most is there's just an attitude that, like, you take on the hardest problems because those are the problems that need to be solved. I mean, case in point is Spot going up and downstairs and opening doors. Like, those are really hard problems. You didn't you didn't have to start there. You could've you could've built that in later. You could've started with a leg robot that just did flat ground and then moved from there.

B I know. But those videos are so cool of the robot walking up the stairs.

C Another story I like to tell is, like, when I was first working there pre COVID and we were traveling around, I came back to the headquarters from, a trip. We were demonstrating the robot to a potential customer. And I was like, yeah. They asked about ladders. Like, Like, they're asking if Spon could climb ladders, and all the engineers were like, oh, no. That's this is impossible. And then we kinda sat there sipping coffee, and there was, like, a weird, like, you know, 3 second silence. And then immediately, people were like, but if we did this, this, and this, like so there's just this, I don't know, insane creative drive where it's like nothing nothing's impossible.

C All we need to know we just need the right signals from the customer that that's actually what drives value. Because the goal is to make these robots commonplace for the good of these industries, to to move them from being things that are seen as r and d projects to operational tools that add value for our customers.

B Maybe to bring the conversation to a close, what do you think architecture can learn from robotics, or what, like, translated between architecture and robotics?

C Your chance to take authorship over the technology that delivers the buildings is to work collaboratively as an industry. So if you're all kind of keeping that to yourselves and not benefiting from that, then the only person serving the industry at large is a software company. And by the way, they're doing all the right things for them. They're they're a business. Right? They're going to they're going to serve the stakeholders, and they're gonna make the decisions that they think are are are best to make. And and maybe those are the right decisions for a lot of architects. So I think, like, it's like a lot of those complaints, you know, I I think they'd go those complaints would be a lot more valid if there was, I think, more knowledge sharing and this kind of collective approach toward these tools. I think that's that's one side of it.

C I'm a little bit out of that game these days, so, you know, not not the expert anymore. But one thing I do notice relative to the type of work I do now is that there's still not there's still not kind of attention to this feedback loop and the idea that my concern, right, is that you're going to see a tool like and this is an open dialogue I have with, like, the people from TestFit. Right? My concern is, like, people are gonna see that and think, like, now I can print a building. Because I can automate the design, that means I can automate the delivery. And, again, it just completely glosses over the, you know, all of the expertise and knowledge that's required downstream from the trades to deliver that building. And it's like, no. You actually what you wanna build in is a certain amount of uncertainty and flexibility for those downstream trades to take in. But what that means is you can't just make a model, deliver a model.

C That means that that model has to stay alive and has to incorporate feedback and update. What I will say is even though that's not necessarily the mentality, the way that these new companies are going around building this software in a kind of modular and cloud distributed way makes it so that they can be like, yeah. Well, if you wanna build that tool, Brian, go ahead. Right? So, like, there are, like, more open frameworks for these things where just because that's not where the effort is going right now and the efforts more on the design side, of course, like, I'm not stopped from doing that. Whereas it would be possible in certain ways, but extremely cumbersome and and probably won't work out too well for me to try to build, like, real time reality capture ingestion systems with, like, what, like Dynamo and Revit. Like, I think I think that probably really wouldn't work out. So I think the emerging cloud platforms are showing the way forward to initially to Band Aid these exist existing large platforms, in terms of building in these feedback loops. And, Yeah.

C I'm I'm excited about a future when when design automation really is able to automate a lot of these things accurately, and we can actually pay more attention to okay. Now what if I take this model into construction administration, or I'm doing a renovation project and I'm tracking it from demo to completion? Can this model live alongside the delivery process and add value there?

B So in Brian's article in architect, he makes the point that most architects tend to see robotics as something very far in the future when in reality, it's here. And what's interesting to me about Brian's career and interest in this is that he's actualizing the implementation to bring automation and robots onto job sites, which is pretty cool.

A Yeah. And granted they're still happening at job sites of a specific scale. Right? We're not talking necessarily about having Spot on your next high end residential project. But it is interesting to see how these 2 worlds are beginning to mix at a variety of different levels. For instance, SCI Arc has a robotics lab that they installed after I left, but it's actually very different from the type of robotics that Boston Dynamics is implementing. And not nearly as cool because those robots aren't inside aren't aren't autonomous. But it's just kind of interesting to see how these 2 worlds are beginning to come together and how not necessarily architecture in this case, but the construction industry is beginning to adopt these newer technologies.

B I agree. And I think, you know, specific to this interview, it's the 1st interview that we're really diving into the disruption on the construction side. So there's a lot of really cool elements to that. You know, the continuous scanning of construction site elements with data dumps at the docking stations, that was interesting. And then, also, Brian walked us through kinda how he sees this opening the door to different variations of automation, new types of robots, and, being able to basically sync data that you're getting from the field with Revit modeling and how the model evolves. But back to his career, I think what's really intriguing to me and the reason I wanted to bring him on the show was just this this idea that he came out of school, and he was somebody who, very early, was passionate about technology, had an interest in robotics, and somehow managed to blend that interest with a career in architecture. And so he's found himself in this unique position, but I think, again, it demonstrates just like some of our prior guests that when you have this, unique passion and you try to find a way to bridge it with architecture, it opens different career opportunities that you may not have known that even existed. You know, I

A think when we graduate from architecture school, we're consistently told there's so many different paths that we can take. And even within the industry, all the paths that come to mind are are very limited and narrow. But this is showing how technology is beginning to expand and open those paths and create those new job opportunities that you're alluding to.

B And I think it's really interesting because it is still an innovative design career path. I mean, working on the robotics side, he's working with the team developing and prototyping this stuff, and then he's working with the different sites, job sites and general contractors that are actually adopting this technology into the field. So he's looking at a lot of different projects in real time and still having that direct connection back to what we would define as traditional architecture, just viewing it in a different way and really looking at the innovative part of how does how does this technology influence the way that we build.

A Right. But also the real time data capture is is something that I think even architects struggle with. Right? So if I am an architect who has been promising my client an as built model at the end of the process, But obviously, the model we have in the computer is not reflective of all the physical constraints that we uncover on the site, or how we had to move a certain aspect to the left or right to really compensate and make things go together. The real time capture means that what we actually hand over at the end of the site is a true as built model. So in the future, I know where all the systems are running, so I can either connect into them properly, or alternatively avoid them as I'm making adjustments to future construction.

B Exactly. And a real world experience that I had, I remember one of our clients wanted that final model after construction. And after we went through the entire construction phase of all the RFIs and submittals that came through, trying to manage that communication between us and the general contractor and just understand exactly where they made their final decisions in the field, it's, like, I don't know. It's a little bit messy. It's hard to completely pinpoint every decision that the contractor and the subs make, that very that deviate from the original model or drawings. And so I found it really cumbersome to track and maybe that's just me, but I think the idea of, like, having tools in the field that when you go out and do a site inspection that you have a way to measure and validate different data in the field and then take it back to the office is seems really valuable.

A And, again, this is happening at a very large scale, and I think some architects might be dismissive of that fact, like, right, I don't work on projects that big, we're never gonna have this type of technology at our fingertips. I think it's, like, important to note that the way technology is ramping up, it can happen within our lifetimes. So even if you kind of set this conversation aside for now, it is worth keeping on in the periphery, knowing that this is something that more firms will eventually be taking advantage of in the near future as opposed to the far future.

B So if you're thinking about a passion that you have that's outside of what you see as the way architects practice, I think Brian's a great example that you can you can run with those ideas and you can experiment and iterate in your career to find and eventually land somewhere where it it aligns and your passion plus architecture equals a great career that's really meaningful. So look to Brian. Look to some of our past episodes on architecture and in this series from season 1 and season 2, and I think you can see the pattern of some of the ways that these individuals are thinking about their careers and taking risks both in their academic experience and in their career pivots to align those 2 things together?

A Absolutely. I feel like one consistent thread with our architecture and people is that they identified actually, I'm incredibly jealous how early on in their career they actually identified what really drove them, and they've created paths for themselves going forward, given what's driving them. And I I only wish that I found that earlier in my career. And I feel like there are individuals out there listening to the podcast who may have found that, but don't know how to to step off of this path that everyone has been put on. But there are are opportunities to do that, and we encourage you to find ways to do so. And on that note, thank you for listening, and tune in next week. Thank you again to our podcast partner, Monograph.

B Learn how Monograph can help you take control of your firm's financial health.

A Follow the link in our show notes or visit practice of architecture.com backslash monograph so that Monograph knows that you heard about them from us. Thanks for joining us on Practice Disrupted, a podcast by Practice of Architecture. Visit us at practice of architecture.com to find out more about future episodes and the changing nature of practice.

B We have several ways you can get involved with our growing community. Find us on social media at practice of arc. You can also become a member of the POA lab or join us on Patreon.

A And if you wanna take your career or practice to the next level, Janine and I also consult, provide workshops, and speak regularly on this research, and we would love an opportunity to collaborate with you.

B This show is part of Gable Media. You can learn more about other podcasts and video channels in our community by visiting gablmedia.com.

A We are also looking for sponsors who want to partner with us in 2021 and beyond. If that's you, please contact me directly at evelyn@practiceofarchitecture.com.

B If you like the research we're doing here, please help us out by leaving a rating or review on Apple. We appreciate you subscribing on your favorite podcast app.

A Don't forget to share with your friends, and feel free to let us know what other topics or speakers you're interested in hearing from. Thanks for listening, and see you next