Transcript
0:00 What if we've been thinking about robotics in construction completely wrong? For decades, we've been trying to automate construction. But what if the real opportunity isn't getting robots to build the way humans do, but fundamentally rethinking how we design and construct buildings? Today on the ATN podcast, I'm joined by Professor Akim Mangis, founder and director of the Institute of Computational Design and Construction, otherwise known as the ICD, at the University of Stoodgar. For almost 20 years, Akim and his team have been pioneers in computational design, robotic fabrication, and experimental architecture. They've been behind creating some of the most extraordinary structures and pavilions that we've seen emerge from this field. We recently had the opportunity to visit ICD in a three-part series to explore their research labs and see some of their incredible work firsthand. But in this conversation, we're going to explore the evolution of computational design, why robotics has struggled to make a meaningful impact in the construction industry, and how AI could finally unlock a new generation of human and machine collaboration. But perhaps most interestingly, we discuss why the biggest barriers to transforming construction might not be the technology itself, but the way our industry actually operates. So, can computation, AI, and robotics fundamentally change the way we build? Let's find out and have a chat with Kim. If you'd like to learn from some of the world's leading architects and design technology experts, head over to ary-tech.network and take a look. And stay tuned. We got plenty more office tours, courses, talks, and events coming soon as we continue to build out the platform. We are live. Akim, welcome to the podcast and thank you so much for coming on.
1:40 >> Thank you for inviting me. Yeah, like I was saying in the the buildup to this, I've always been a huge fan of the work you guys have done. I mean, from the computational parametric world, whatever you want to want to brand it, your you know, the pavilions and the stuff that you guys have built out of the ICD department has been iconic. So, I think coming to tour the place and see the models and how it's all done was was really inspiring. So, thank you so much for joining. Also, as a a buildup, anyone that's listening to this, we've also done a tour around ICT ICD department and you can go and see all the different work that they're doing.
2:22 We did kind of three little mini episodes. The podcast is in the middle so you can go and check them out and get the full spectrum of what we're going to talk about. But of course, we want to have you on on the podcast to talk about really the founding of how ICD started. you are the founding director. So talk us through that story. How did ID start? it's almost 20 years old at this point. So you know you're a pioneer in this space. Talk us through how it began.
2:54 >> Okay. as you mentioned is almost two decades ago but I kind of vividly remember the day I was made aware of the call for the new professorship at the University of which for me was a very good match because the for romantic reasons or not but the only school that I would have considered to return to to Germany where I was actually born after spending a lot years in England at the the architectural association would have been the University of Stat primarily because it was aligned with my research interest in those days. I mean I'm sort of the generation of digital architecture. It's not really the origin of when digital technologies entered architecture in the late 60s, early '7s, but when it hit sort of architectural practice for the first time and what what always surprised me is somehow the rupture between the digital we call it digital at that time digital design turning into buildings and how uneasy this transition was. Almost as if you digitally constructed an enormous problem for yourself that you then had to both rationalize into something that was partially constructible and a lot of the qualities that were so striking on the screen very often got somehow lost in that transition. So I was since the days that I studied at EA looking for alternative models and of course sooner or later you come across pioneers like Fryoto who had device design methods like his form finding methods that were non computer computational I would say but fully aligned with the subsequent materialization and I was always wondering how can we actually use dig technologies not as something that separates between the virtual and the physical world but rather something that allows us to integrate let's say the the the logics and constraints of materialization into the design process from the start similarly as Frank Otto had done with his what you may call analog computers the models that he built so and that's why I I was actually very tempted with this call from the University of Stoodgat and then it happened that I got the position and with that somehow also the agenda was set and I think what is really striking is that this culture that I expected to be somehow alive at the University of Stoodkat people being interested in construction people being interested in materialization as part of let's say the cultural project of architecture.
6:05 and this being founded in an interdisiplinary approach where there's no where people from engineering, science, architecture, natural science all work together on eye level that this really was was actually there. so it was could have been a very I don't know nostalgic path to disappointment but it turned out to actually meet all expectations and surpass them.
6:37 >> Yeah. And so you founded it's IC ICD institute of computational design but now it's the int institute of computational design and construction right did that element come a little bit later or you know initially it was it was computation design and then you expanded >> yeah of course very important question and a very sharp observation on your side that the name actually changed in 2014 mainly because we realize okay there's this idea that we inform design by materialization but what happens if we also design the materialization processes themselves meaning that we can actually utilize the big difference between robots and actually dedicated CNC machinery where I would say okay a CNC mill still does milling CNC saw still does sewing a robot is actually an entity that you can design tasks for. And that expansion actually led to more profound interest into all kinds of digital materialization processes. Most of them then turned out to be let's say revolving around initially industrial robots and also other kinds of robots.
8:03 and thus sort of that was formalized with adding construction to the to the name of the institute because in the end I think computation is the means to integrate design and construction and that's has been the ambition from the beginning but it got more and more really a point of convergence and that's why we paid justice to that with actually amending the institute. >> Yeah, it almost seems like a a natural evolution of the the technology and the process. But maybe paint a picture when when you first started the institute.
8:44 What was the landscape like? So it was it was kind of you know late teeny 2008 2009 Grasshopper was around and you know but then you had the kind of classics of I was forget was called design systems or what was the landscape when you created it? >> So the computational design landscape was of course very different from today. Rasop actually wasn't around at that time. I think David Rotten sort of still worked on it if I remember it correctly.
9:18 So the first couple of years we actually programmed everything in Rhino in Visual Basics and then so it was a different and more let's say just everything was harder. I think what was around was the whole idea of parametric thinking. So I was lucky enough to join the old smart geometry effort that started in the early 2000s. we also ran various workshops with this community at the architectural association at the time.
10:03 So, Generative Components was of course I would say the predecessor to Grasshopper >> I was just trying to remember the name Generative Components that was the the OG Grasshopper basically >> and and and what I think is a quite amazing to see is that this whole generative components community grow into very interesting personalities around the globe. So, but that was the landscape. I think we saw the first robots at for example the thuric I think bought their robot a bit earlier than us and they did some really interesting work but of course it was by no means any mature research field.
10:51 I think a handful of people that each that we all know knew each other sort of embarked on this journey through this various activities in the early 2000s and then sort of it it got a more and more widespread interest and received more and more widespread interest. >> Yeah, I'm also was a huge fan of smart geometry. I I often talk about it as that was one of the most impactful events I went to as as like a young architect interested in this world of computational design and and it's really partly why I built ATN was to kind of build on this kind of community building that you know essentially smart geometry and other platforms laid laid the foundations for. So so yeah I think that was a a really amazing moment. I think one thing that's that's really interesting about you guys and you also mentioned ET ETH as well is I think architecture we do have this issue with academia and practice and and traditionally they seem to be growing further apart but I feel like your research and the way that you guys do it it's it's very you know applicable or very close to being kind of real world you know usage or Some people take the technology and create startups and things like that. Talk about h how have you kind of tried to maintain this real as opposed to becoming too engrossed in the academic and you know growing away from real world practice.
12:33 >> Yes. So what I think maybe is interesting that these things things are related. I think when we started actually what we do what we were doing was not even taken seriously as a kind of academic endeavor. So there was no research funding there were no research bodies that would have reviewers to assess any grant application that you would send them. So it was actually very immature field and then I think also because of this great community by no means I think that something is limited to ICD. I think the guys at ETH, Matias and Fabio and also Philip Block that contributed majorly to this that we were taken serious as researchers and that of course there would be a facet of architectural research that was fundable in a sense and that there were also a critical mass of reviewers eventually that would understand it and young researchers that would build an academic career from it. all the things that somehow indicate that a field gains somehow academic traction.
13:43 But with that of course also comes the responsibility that you do something that society benefits from probably a bit more directly than just on a sort of larger speculative cultural horizon. And that's also the expectation of the funding funding bodies that we're very privileged to have in central Europe that actually fund this kind of research. And I think what was very interesting to see is that in the end we didn't have to compromise on the methods that we were interested in. The computational design methods I think are still very speculative, very experimental at times. but it was very interesting to see that you can apply them to I would say experimental endeavors that also in the end lead to a very experimental architecture but you could also make contributions to architecture and construction that are more tangible and implementable in the short term. And I think this is a was also a very rewarding part of the journey. and and quite frankly it made the journey also very interesting because all of a sudden you face the interesting situation of all the certifications that you need to actually see an experimental architecture through that. Initially it seems almost un impossible to actually build something that is beyond what is currently done and then you figure out slowly how to do this. So I think it's also a very rewarding part of the journey to move more and more towards actually implementing the methods that we developed and the technologies into actual building projects.
15:41 Yeah, I think that's that's maybe the key part is building stuff, right? When you when you build stuff, you it suddenly becomes very unacademic. You have to get this piece of wood stuck to another piece of wood and you know it's it's swollen and you know you have to solve these real world problems. So maybe that's the kind of beautiful link is is you guys are famous for building all these amazing structures and pavilions and and that's what links you to the real world I guess.
16:12 Yeah, I think it's also increasingly that's the part that is less visible but also all the contractual frameworks in which you need to operate all the let's say management that is required to run projects that do things out of the box navigating I think also the institutional boundaries that we have what you're supposed to do as an architect and what you're supposed to do as an educator and all these things. I think a lot of a lot of challenges that are probably less prominent in if you just see the the the work but are equally I would say exciting but also time consuming and challenging. Yeah.
16:59 >> Yeah. Yeah. Yeah. That's always the fun part. How have you seen I mean from you've kind of had this department during the kind of boom of computational design or you know at some point it was like the parametric era right how have you seen it evolve to what it is today I mean you you founded it when grasshopper didn't really exist or it was in its very early days then we had this kind of parametric era and now of course this boom of AI and architects being able to kind of vi code their own you know scripts and stuff like that. Has it has it has it exceeded expectations? Where do you feel we are now in in the world of computational design?
17:45 I think from the start I think there were two parallel streams no in the way you could employ digital technologies and I think that's still kind of present what that's why I'm getting into this but I think there's a way of employing digital technologies to enhance what you have always done make it a bit more productive make the processes a bit more efficient or you use digital techn technologies to rethink design and construction from its from their foundations and I think we've always been interested in the latter part but the former part was also always there you know so I think and I think what we have seen is that also of course employing digital technologies for real world problems making real world processes more efficient is something where you can have the impact much faster faster and that's also why that gained a lot more traction and I would say also succeeded in many ways because all design nowadays is produced digitally you know not all capitalizes on the all designs capitalize on the fact that they are digitally conceived but and that's what we are interested in so how does how do digital technologies open up new opportunities and usually for those opportunities to be really interesting you have to consider design and construction in an integrated manner and of course with AI methods over the last 5 years the a bit longer actually but the the possibilities have been exploding.
19:35 >> Yeah. that that is very clear but I think still most of I would say the AI progress in practice we see is making relatively mundane tasks seriously more effective let's say >> yeah yeah I think that's a good point it's a nice way to put it two different streams is it's either it vastly increase your productivity or it rethinks maybe either the design process and the design and build process which is of course I always think as the holy the holy grail where you're kind of designing it digitally and and building it digitally but do you think practices have really embraced the true potential of computational design at this stage >> I think from the perspective of really achieving a high level of integration of design and construction and that by that actually rethinking construction not really I think there are a few examples out there and of course that's also our main research domain so we really try to push that but I think that hasn't really fully been embraced and with that I would also say that the full potential of digital technologies still will still needs to be discovered in construction but it's also understandable why not because it challenges it as I said before it challenges institutional boundaries it challenges business models and it also challenges business models of architectural practices and contractors. it it actually challenges the way we currently deliver design and construction and that will take a bit more time. So we are very happy that we have this few opportun have managed to let's say create these few opportunities where we can actually showcase where this could go if you if you are in a position to implement this high level or or actually really enable that higher level of integration.
21:42 Do you see the the ultimate adoption would be the architect being the master builder, the designer, the builder, the deliverer? Would would you say? and I I I think what would be the ultimate goal is a real shift in let's say designing construction culture where there's a spirit of more integrated design and delivery and I think this is not the case with current business model of design firms and of construction firms.
22:24 Yeah. Yeah. So across the board >> I mean there is something like design and build of course especially in the not so much in central Europe but in the UK and the US but that has more or less is the single goal of making construction cheaper. So I think we need to have something that you cannot just measure in euros per square meter but they should also add architectural quality and opportunity and possibilities.
22:56 >> Yeah. Let's touch let's go back to the the introduction of robotics. So you're saying in the kind of you know 2014 was it you you kind of added the name. talk us through how well adding robotics it seemed like a natural fit, right? You you were designing the things and then you have to build these these kind of elements. But was there a a moment that made it more accessible in terms of being able to access that technology at that time?
23:29 >> So I think the we got our first industrial robot in 2009 I think. that was a secondhand robot which was very accessible from a kind of cost point of view but extremely inaccessible from a a skills point of view that you need in order to actually operate the robot. I think the robots are still somehow same costs. now we have also done a lot of research on how you build a lot cheaper robots and probably more of them and how you can make them collaborate. So there are different avenues of getting into construction robots. The real shift I think is now what is happening with AI is sort of the lowering the skill barrier that was required until very recently to operate these machines in a meaningful manner. So that that is a huge a huge change in human machine cooperation. and I think that opens up a lot of new opportunities and if you combine that with sort of what is yet to come with spatial AI and then also ultimately with physical AI is a lot to be still discovered and expected in the future.
24:56 Yeah, I was curious actually something that sprung into my mind last night was and when we were in Stoodgart was has there been any connection between you and the automotive world because you know for listeners that maybe haven't been to Stoodgart don't know it's the automotive hub right so you must have robot you must be surrounded by robots making cars and things like that has that had any crossover >> at the very beginning actually we had a number of projects from with Mercedes-Benz and interestingly enough they came to us and they said we have a problem they want to have a much higher level of individualization in our cars and that's a problem that architecture resolved because all buildings are different okay that's interesting I we always looking up to the automotive industry as sort of the high-tech industry and of course a industry is being a bit more on the other end of the spectrum. But then we worked with them for actually several years and we had PhD researchers actually moving to Mercedes and also moving to BMW. And I think the the the situation or the context is still so different that I think there are certain things that you can learn but in the end the possibility for direct transfer is very limited and I think that's also what we experience you know when we we have now very I would say institutionalized collaborations with our colleagues from production engineer ing from mechanical engineering and initially they say we've all figured this out for the automotive world 20 years ago but then if you really get deeper into it and you see that let's say the approaches and the related or the required methods are so different you realize that architecture engineering construction pose different challenges and I think that's that that's that's important so it's not it's not a simple transfer there are things that can be transferred But u at the end you become to realize that architecture is just fundamentally different from building consumer products.
27:16 >> Yeah. Yeah. I guess the repeat repeatability element they're just building essentially the same thing many many times. >> well I think also the biggest difference sorry I didn't want to interrupt but the biggest difference is they can run 40 50 sometimes hundreds of prototypes and physically test them to destruction before they go into the actual production. >> Right? So we build and that's that's mainly also what if you get a bit more let's say daring with the actual loadbearing parts of of an building what we are doing with our colleagues from structural engineering you cannot do this with the building so you have to make sure that what you build is sound and safe >> without actually destroying it 50 times before ends and that is the key key difference know the mode of verification and the required design engineering methods that's why they differ profoundly.
28:20 >> Yeah. Yeah. It's difficult to test the buildings in in reality before you build them. But I mean the the construction industry has always had this obsession or this fantasy with bringing robotics into the industry. And it's fair to say it hasn't been so successful yet, right? You know, famously Japan tried to bring in robotics and you know even here in the UK we went down this road of prefab. Now I actually think one of the mistakes was there was not much robotics in these giant factories. It was just people putting inside them. But why do you think the construction industry has struggled to to integrate robotics in a meaningful way and why is now the moment where we might see this actually integrate?
29:13 I think what we've seen over the last 15 years is that if you just employ robotics to automate previously manual construction processes, it is not very successful quite frankly. because construction has evolved over hundreds if not thousands of years and all those manual construction tasks are actually very effective.
29:45 So if you just employ a robot to build a building as you knew it from manual construction, I think it's it's very difficult to actually compete with these highly evolved processes. And I think that's also what happened partially in the UK. I think you only have you can really only leverage robotic construction if you're ready and willing to rethink your actual construction methods and the related building systems, structural systems with the opportunities that offer that are offered by robots because robots are good at building buildings different in different ways than humans. so I think that's the important part and that's the kind of higher level integration that we see now. So you rethink design and through rethinking construction so that you design buildings that are built for teams of humans and robots to construct them together because I also don't believe in full as I say full robotic construction. I think that's also somehow a misguided vision.
30:59 I think there will be always humans involved and I think it's about ways of skill sharing. and also and that's the key part in my eyes developing new construction methods and new building systems that are designed for this new situation. >> Yeah. >> Yeah. We call these other approaches pre-digital. No. to use a digital technology like a robot to implement the pre-digital building and that's I don't know I think what we've seen in in last few years has been demonstrated multiple times that that is not a viable way of going about it. So like a really dumb example would be a robot laying bricks, right? It's it's kind of like the brick is very human. It's it's literally so that we can do it and and a robot just replacing the human to make a a brick wall versus would you say 3D printing where it's kind of, you know, the the machine is new, the robot is new, and also the way that we build is new. It's it's kind of layers of concrete. It creates this whole new world of opportunity. Would those be two good examples?
32:13 >> Would be a good example if we also add in the environmental challenge that we are facing which we need to tackle at the same time. So I think it's those two things no digital technologies and I don't know addressing the climate crisis and u increasing the share of biogenic building materials that maybe 3D printing is not the very best example but I think even even in the work we do on timber construction you can see that there's still a lot of opportunities for rethinking timber construction through the lens of digital technologies.
32:57 >> Yeah, I think that's that's a perfect segue to to talk about that because a lot of your work does or a lot of the research you guys are doing revolves around timber and robotics. There's also a lot of biiobased materials. so you know a robot CNCing some kind of timber building element and it gets assembled into some kind of structure. but what are your thoughts on on that compared to 3D printing in general? Because I've also seen you guys do 3D print elements of the building, but I don't think I've seen from you guys this notion of you press print and the, you know, the machine just prints the building. Where does your view sit kind of robotic timber approach versus 3D printing?
33:46 Well, I think because from the beginning we have been more interested in biogenic materials. So primarily timber and natural fibers. and because the sort of large scale material at a large scale construction scale 3D printing mainly revolved around mineral materials concrete with a very high cement ratio. we were not interested in that. We did a lot of 3D additive manufacturing for responsive elements.
34:24 We also have done large scale additive manufacturing for reinforcement of timber elements. So we've done a lot of work in in larger scale 3D printing but not what is now I would say also already fairly advanced technology the kind of layer printing of concrete or or now also earth which I think is is very interesting and is definitely those are technologies that also force you to rethink how you deliver buildings. I think they need to possibly foreground a bit more the environmental side.
35:04 but generally this is something that I think u is is also very interesting topic is something that I think we did a bit of take a bit of different trajectory simply through the materials that we were interested in. >> Yeah. Is there any startups in this in the AEC world out there that you think are quite interesting in terms of the technology they're using that are you know a linked align to your own kind of views?
35:37 Definitely I think what we now see currently is that there's actually a good amount of startups most fall into the former direction where I think what I what we discussed beforehands that they find very smart ways of for example employing AI methods to automate automate parts of the design or construction process where you can have immediate productivity gains. I I appreciate those startups that go the sort of hard way of also engaging with the hardware of the building which is a lot tougher and I think there are many interestingly enough also especially in Europe many interesting not many but there are a few interesting startups that actually go down this route remains to be seen and how far that goes but simply let's say the prospect of having an alternative building industry that operates from the ground up on those principles and has related business models gives me a lot of hope for eventually coming to this higher level of integration that we're seeking.
36:56 >> Yeah. Yeah. I mean obviously speaking for myself I was part of Facet Homes which was we were trying to go down the route you were saying like we were trying to reinvision the entire process right we were we were kind of like our own master builder it slightly more simplified approach where it wasn't a robot arm but we had a CNC machine that that we could fabricate our own building system that was different to the traditional building system. Whereas you have other companies that yeah are trying to automate that process. Like automated architecture is a good example where they're trying to automate just the timber frame. I think that to me seems to be the two ways I see startups entering the space. you either automate an existing process and maybe you use that to to get your foot in the door and then expand you know what they can do or I guess icon is is a is another example where you try and tackle the whole thing which is is very difficult you know you're you're almost getting into a developer if if you go all the way but yeah I see a lot of startups those are the two ways I see it and I think you know both are very challenging But it seems like the way to start is to automate something and try to expand from there.
38:23 >> Yeah, I think for easy market entrance, you definitely have to go and initially take the smaller step of automate something that's where you basically fill a gap in the larger construction chain and then try to expand on that. I I do do agree. Share your observation. Absolutely. But I think the real gain is in a truly integrated >> same approach. Yeah. >> Well, like the the vertically integrated approach from design to construction.
38:56 Let's let's touch on AI at this at this stage because I feel whenever I talk to people like yourselves and and people that have been working in this space with computation and robotics, AI seems to be this kind of new, you know, it's almost like this new drug that gets integrated into the robots and it almost seems like that it's going to enable this technology to to kind of get to the next level, get integrated into the industry. But what has what have you seen AI really unlock in terms of your own research both on the design side and also the robotic side?
39:40 I pick up on where we discussed beforehands. I think what AI definitely has enabled and that may be good or maybe bad but it definitely broadens let's say the the applicability is that lower the skill threshold that is required to for example operate robots and I think this is somewhat important because in my experience the barrier for adopting robotic fabrication for example in small medium scale businesses was less the investment cost more the fear of having capable and skilled labor that can actually operate such an advanced manufacturing technology in a sustainable manner. So cannot have only one guy and the guy gets sick and then your whole production comes to halt. and this was always a huge worry and I think that's of course can be addressed with AI enhanced modes of human machine collaboration. I think there will be also a real push in how determinant your manufacturing processes need to be. So that will liberate a lot of potentials for more effective production from our own point of view. If you don't have to, for example, specifically determine in our, for example, the the fiber winding process that we do the exact geometric position of all the points that the robot has to travel through, but the robot can actually find them themselves. You basically strip out 3/4 of the complexity of actually programming machine to do it. and you also lower the error margin quite a lot. So I think there's a there there's there's real gain and I think once we talk about more distributed robotics so beyond the factory but onsite construction renovation physical AI will of course give it a huge push also that you can imagine that if you shift from this idea that you need to have a world model to operate in to actually even get your robots run and set up and you can have very local orientation, local intervention as actually CR people do in a renovation site. No, they don't they don't know the whole building. They just see the problem and they fix it. and that in a very autonomous manner, I think that will open a lot of new opportunities. I think then we also probably begin to talk about machines that are really have a high level of autonomy which means that they can interact with fairly autonomous humans on a totally different level. So I think we can also expect a different way of human machine collaboration.
42:42 so on the robot side I think there's there will be super interesting opportunities that will allow us to further shift let's say I mean the effort from just consuming materials because it's cheap and easy to design and construct to actually getting really fine grain use highly differentiated structures and they do exactly what they're supposed to do without an excess material which we still have in construction everywhere now because it's just cheaper to pour some material in somewhere than thinking about it whether you really need it or not. So I think that's that's great. I think in terms of computation or let's say it's in terms of the contribution of AI methods to design sure there will be a lot of or there is already fantastic advances in all kinds of datadriven simulation approaches where you can just strip out also a lot of complexity and and also a time and you gain a lot of speed so that you can monitor also your design decision making much much better.
44:08 when it comes to let's say spatial generative design probably remains needs to we'll have to see where that where that whether that really is going to fulfill the huge promises that are out there. Yeah. I'm not I think I'm not pessimistic. I'm also optimistic in this regard. But I think it will not be a little effort to come to something meaningful to come to a point where true and generative AI design actually generates architecturally, constructionally and environmentally interesting solutions, not just something. Yeah.
44:53 >> Yeah. Yeah. It almost seems to go back to the AI's injection into robotics. It's kind of giving robotics a bit more of that brain, right? So, whenever I talk to people that work with with the robots, it's it's very deterministic. Like, you pick up this thing and move it there and it's very accurate. And if that thing is moved slightly, you know, the robot can adapt. And it seems like AI is is giving it that brain to be a little more autonomous. And maybe even there's a future that people can now just just type and talk to it about what it's got to do rather than knowing the code and and that kind of unlocks that accessibility to to so many more people.
45:33 >> Yeah, I think those games are mainly in the computer vision part now where actually there's no computer vision without AI anymore. >> Yeah. So I think that's let's say what you I think is adding decent eyes and brain to the robot. That's that's definitely a huge leverage. And then I think of course the subsequent parts of having a higher level of different means or multimodal collaboration is of course also a a huge game. I think what what of course interests us the most is not only how do we do that but what kind of better let's say architectural solution can we generate from it.
46:28 >> Yeah. I think I I always get sent these like memes or videos of of like you know maybe one of your robots or another one that's like picking up a piece of wood and building something with it. And if you showed someone that 5 years ago, they would be like, "Oh, that's cool. Coding and robotics." And if you showed them the exact same thing today, they're like, "AI is coming to take over the world and take take over your jobs." And you know, as soon as people see a robot, they now think AI when when typically these things are just kind of following quite quite strict instructions. But yeah, I'm sure that's one you've you've come across. But what do you think it would take or what do you think the timeline of you know real disruption in the construction industry like what does the construction industry do you think look like in 5 years or or even 10 years will we really see more integration of robotics in a meaningful way?
47:28 >> That's a really tricky question. Yeah, it's a mill million dollar question but >> I think it really depends on the success of construction startups quite honestly because I think the construction industry as we know it is actually very slow and also very reluctant in adopting a lot of these technologies because as I mentioned before they partially also challenge their own business models. so you know when when we started we thought okay the industry is waiting for us to sort of show them that can be done and they will just adopt it but of course that's not the case.
48:12 >> Yeah. >> So I think that was also one of the incentives in actually getting into the building of real buildings because we wanted to demonstrate that it's possible. No, but even now that we're doing this, it's not that let's say all stakeholders in the construction industry find this particularly attractive or something that they want to embrace. So I think we need first we need a different culture in construction where you basically have also different mechanisms of risk takingaking.
48:53 And second, we definitely need to grow an apparel construction industry with startups that really embrace those technologies and understand them as an opportunity, not as a threat first and foremost. >> Yeah. >> Does that change what you guys do or the end goal for you guys in a way? Because another way to look at what you guys do is you could you're an incubator for people to take this research and I'm sure you know many of of your researchers have have probably done this but your research the research evolves into a into a startup.
49:33 >> Yeah. I mean we have a couple of spin-offs and startups from from the institute. I think we also need to be a bit careful with the term startup because I think there's a difference between a startup and actually founding and running a construction or a company. So that these things have different purposes usually. but yes that's definitely an ambition and generally I think I think the greatest accomplishment of ICD is not the pavilions that we build or now the buildings that we build or the building systems that we develop but it's of course the people that we had the privilege to educate and to train and I think they will also have by far greater impact in the world than let's say building another tower somewhere for constructing another building and I think that's and what what is really great to see is that those buildings and those people our researchers not only venture into academia but they actually a large proportion also goes into the building industry and I think they will make the change eventually. Yeah, that's a great point and that's a that's a great way to maybe ask one of our the questions we like to ask guests is say someone is is watching this and they want to come and be a part of ICD and and and join your department. What do you usually look for? Do do they do they have to meet a certain criteria? What kind of people do you look for when bringing people in?
51:12 Yeah, we don't have a fixed profile or let's say very specific aspects. I mean there's let's say the the requirements of being skilled when it comes to coding and programming of course previous experience with digital fabrication also does not hurt but in the end it's about finding people who are dedicated and passionate to rethink conventions and actually really fully embrace in critical manner the opportunities that dig technologies offer and I'm always astounded I mean just today I went through interviews for a new job post that we have and we received 350 applications and there seem to be a lot of people out there who are very interested in this line of work >> and that that also I think >> somewhat confirms that there's how should I say this this cultural change I think is not u just a kind of speculative goal but something that you can see on the horizon. Yeah, it it doesn't surprise me because I I'd imagine you have a lot of people interested to come come join your your crew and yeah, if you ever need us to help you spread the word, we we can also post it on ATN and get you even more applications. But yeah, it's a very cool thing you guys are doing. Thank you so much for hosting us. Like I said, if you want to find out more, you can see the episodes where we actually tour your studio. We also went to your robotics lab. Actually, maybe a finisher, you can explain to people that you you guys are building your own, you know, robotics lab, right? Which is under construction or or or about to go into construction?
53:12 >> No, it's under construction. So it will be the future home actually climate neutral home of the cluster of excellence in CDC where ICD is a crucial or essential part of and yes we take it very seriously when we say we want to rethink design and construction. So we thought the consequence of that should be that if we get a new building generously supported by the university of Studkat we should really make sure that it physically embodies our research and all the loadbearing components in this building are the outcome of the research that we do. we also implemented with off-site robotics and on-site robotics through actually all the technologies and also the machines that we developed in house. So it's that's a major milestone for us hopefully to be completed in late 2028 early 2029.
54:16 >> Yeah. Amazing. Well, it's a good excuse for us to come back and film the finished piece or we can host an event there or something. >> That would be great. Yeah. >> Well, thank you thank you so much for having us. Thank you so much for coming on the podcast and telling the story of ICD and and all the amazing work you do. And thank you for inspiring me as as I was really interested in this space and always looked at what you guys did with with admiration. So, thank you.
54:48 >> Thank you for having us. It was a great pleasure. >> Yeah, we we'll be continuing a kind of construction theme. I think we're going to after we went to you, we've we've also done a tour of automated architecture. So, guys, stay tuned for that. We'll put this up in the next few weeks. Otherwise, Kim, thank you so much and we'll see you guys in the next episode. >> Thank you.