EP 4 | ArchiTech Office Tours | Foster + Partners

what's up guys and welcome back to the architect Network and today we're back with our architect tours on a beautiful London day as you can see and we're the one and only Foster and partners here in London now this is their office but it can actually be described more as a small village a lot of these buildings around are all part of The Fosters complex we're going to be going in to talk to the Ard team and listening about some of the technology that these guys are working on but I'm so let's stop get inside and check it out [Applause] [Music] all right guys we're here at Foster and partners with the one and only Marius from the Ard team here um Maris is one of the partners in the team maybe you want to introduce yourself and the the team little B uh first of all welcome for partners uh my name is marus I'm a partner of the applied R&D team um so today hopefully we're going to show you around our uh campus and uh talk about technology I suppose or any other topics that ol comes up with uh so maybe we should head on yeah let's go so now we're heading to the main studio so this is the I think it was the first part of our campus here in battery and still part of the campus uh and this uh staircase leads to our open uh plan uh like office desk that epic walkway yeah it's quite quite Grand day we have a small Cafe here we can get food and beverages this is this is the bloom headquarters you maps in the building in the city very nice building [Music] incredible uh facade amazing buag technology super green building right yeah [Music] [Music] so this is the main Studio this is where all the central design teams teams that are supporting the architectural teams are located and anything from design communication visualization graphic design our very own applied R&D team yeah F down at the end and obviously yeah it is an open space we have meetings on the side table it's very integrated you can go to someone's desk for example you can go to the structural engineering team and just ask them something it's just next to each other it's actually super nice having that breakout space because I think we always get like jump into a meeting room they're all books it's really nice to have that kind of breakout space okay let's go to the side yeah so I head up the film and visualization team within the practice so we have uh 35 people within my team uh all the production team set up in the main studio all the 3D artists so down here what you can see down here is the editorial team so down here we have four editors and uh two cameramen so essentially once all the 3D content's done so we can do anything from live action film so filming our sort of projects whether post docy or on site or or we can do a lot of work where we do competition films so it's working with the architect teams um trying to do pitches for new work essentially so it's really becomes part of the brief uh briefing documents now you need to do a film for uh these projects pretty much most of the key ones yeah is there any particular software that you guys are into right now for example unreal is that kind of maybe at the core of what you guys do or yeah so I mean again like a lot of fir you know we we'll use lots of DCC packages 3s Max Meer um but we have a so a team dedicated to using that sort of those sort of DCC tools uh we use nuke for compositing up there um and then we have a separate team uh using Unreal so our real time team so we bought the games engine when we did Apple campus we put the whole of the Apple campus in a games engine so that's when we did it a long time ago was cry engine at the time uh but then we' switched over to Unreal since um and then we have another team which is Motion Graphics so again communicating these complicated designs and elements of the complicated designs is really tricky so you know we need to do it so a lay person can actually understand the scheme so the motion graphic plays a key part in that that side of it basically this is our sketch model shop so it's a place where every designer in the practice every architect from architectural assistant to have the partner level I suppose can come in and they can prototype physical models the laser cutters and there's yeah an abundance of uh apparatus from uh laser cutters zo machines anything down to simple uh foam making Cutters again a lot of uh materials at the disposal of of the designers so if you want to mock up something for your design proposal you just come down here you make it you see how it looks and you propose it is being reviewed so the normal design process that you would do in the University still exists here yeah um I think this is like where you you're christened into the design process right we often have like a lot of Das design assistance Splat ones come in here and I kind of get the sense it's part of that beginning process of joining yes yeah and physical models uh it is true there are essential even though we have so many visualization tools and technology has progressed so much there's still like something that is tangible you can place it on table immediately you can even assess areas if you want to like if you color code blocks for instance so it's a very quick way to provide this uh intuition and this feedback to to your design I suppose so that's why we have all um this big space for our designers and also like there members that of the support teams that are professional model makers are here to assist with anything that any complicated tasks let's say or any specific uh needs or to recommend materials that people can use for specific designs this is where we have all our uh 3D prins wow yeah these are serious like the powder printers right yeah these are po printers and they're pretty in color as well and we have nylon printers here they're yeah most of the time they work the work full on like overnight uh to deliver designs that can be reviewed the next morning so do you have more of the ultim makers like the desktop ones as well or there I've seen a few flying around uh but um they're kind of hard to maintain uh so it's better I mean since you have these facilities here you don't have to deal with uh preparing the files yeah yeah to be ready to live be U 3D printed yourself so this is handled by the professionals I suppose so you don't get away from this process of actually preparing the file yourself these do look like big big boy princes these ones blue foam which is now this is also it's uh yeah I don't think we ever used blue foam it's always a white foam I have to say uh I haven't been here uh this again many times but I never saw blue to be honest uh we prefer white as a practice I suppose it's definitely another uh christening of every architect is to go through the blue foam or white foam modeling process you so you guys mostly print with the powder printers rather than a mixture so we've got P we've got two big SLS machines yeah print with like a nylon powder then we've got the color jet printers that print with a gypson powder okay so all of the color models are on these ones yeah so they're quite fragile but they're good for like massing and they're quite quick so we can fabricate stuff for the architect like next day yeah yeah and then we've got three smaller machines like a smaller build volume with a resin so that takes paint really nicely and you can get super fine with these ones fine like down to4 and then you have a lattice support structure so this is all sacrificial we would we would get rid of that and that's what's attached to the platform so it grows off the base and then you remove all of this Lattis work and you're left with your par but you do see a witness where the nodes see they tap it down little dots yeah like a little point and it just tickles off basically so this will be Marius you you're scripting this part yeah you send it to you that's um that's the workflow that's the L St yeah yeah [Music] [Music] so this place is first and foremost a materials Library uh and we have over 25k samples and then and I think the purpose of this place is to provide a knowledge base for the practice on materials products um technology sustainability you name it processing and we could get asked anything on a regular basis um The Architects will come along say we're working whether it's a competition or it's a fully-fledged you know project what's local materials yeah what can we look at and then what are the craftsmanship the technology IES what can we do with it at the end of life and you can see here yeah um you know circularity is a very important facet you know and what can we look at but we're lucky that we got an awful lot of Architects and companies and you know all working with us to try and push things uh and that's what it's about I have to say this is one of the most impressive material lab labs like I've ever seen but it is like I said you know because we have a range of people and I have a I've got such a plethora of uh knowledge and expertise so it's important to stress that um and we try and do sort of uh cpds exhibitions as you can see but it's all at the Forefront so what you're seeing there is caline Clay which is the latest in low embodied carbon concrete but what that's what I say people coming to us but we're also going out there to say a decarbonization you know what is happening in concrete in glass in steel in aluminium and that's all of the work that my team and I do my name is Martha ziger I'm the head of the applied research and development group at Foster and partners I'm a senior partner at the office and have been here for 18 years mafa thank you so much for having us uh you know the tour was amazing I'm always blown away by this Village that you have here and like some of the facilities you have yeah is uh yeah it's mind-blowing maybe start by introducing the Ard team um at Foster partners and what the kind of disciplines that you currently cover in the team before I started with the applied R&D it is important to understand how we work as an office so um the office is an integrated practice which means it has a lot of different teams that work together from conception to completion everything in between so we have our own Engineers me our own product design uh people who do film and visualization have you seen Urban Design landscape design and within that spec we also have uh the applied research and development group our group has historically been uh very small uh we are often 1% of the entire company so now that the company's around 2,000 people we're around 24 yeah and it's a very interesting mix uh because all of our backgrounds um most of our backgrounds in the past was Architects engineers and our common denominator was that we knew how to program we had some sort of computer science degree one way or another uh but now we're expanding our group to include actual software developers backend frontend full stack data scientists but we do have the old artists for example uh we like to mix and match and we like to bring a lot of passion and different disciplines together in order to be able to solve uh interesting problems in fascinating ways one of the thing that is important also for us to understand is that we perform in different areas of expertise so our bread and butter 15 years ago was complex geometry and that's how we started with complex geometry and Fabrication with parametric design back in the day but obviously as technology evolves we evolve too and that's also part of our remit we need to see far in the future to see what is going to come 5 10 years from now and not necessarily just within our industry but any other industry outside our own so now we're focusing a lot of different things with a performance driven design optimization writing our own analytical tools that run in GPU for example and creating our interactive application that allow for realtime performance rent design uh to to be available to uh the designers in the studios we are doing a lot of things around inter parability so uh connecting different uh software together in real time for people all over the world world we have been uh focusing quite a lot on AI and machine learning for the past 5 six years developing applications for the office uh that are going far beyond gen which is what everybody's speaking about right now yeah a lot of Investigations uh around extended reality so augmented in virtual reality and that's has taken us always almost uh for my entire career here we have been investigating what is the newest headset and what are the applications that we could uh do in relation to what we were seeing was coming towards in the future uh and we have been doing a lot of things around digital twins smart cities and smart buildings having developed our own iot system robotics we were the first architectural practice that had spot in house and used it on a proof of concept yeah um yeah that's a little bit of too yeah you're really so yeah you're covering a lot of disciplin uh which is super interesting uh maybe one question like how do you one problem I always had is there's so much going on right now how do you know what to focus on or how do you prioritize some of these things like AI is popping up or you know there's new things in AR or VR like how do you determine what to follow it is part of the job description right to be able to identify these things and you need some experience and there I say some Talent I'm not talking about myself I'm talking about an entire team behind me who's supporting that and they're all extremely talented people um and we we are the the interesting thing about people in uh applied R&D is that they have a lot of different interests and they're looking into a lot of different things all the time so we have an overall strategy about things that we see and because we are in different domains we know that they're going to pick up uh but also we get different ideas from people coming in and suggesting oh maybe we should be looking at that and the main thing what I think makes us really powerful in that respect is that we are looking outside our industry first and foremost when we started looking into AI back in 2017 we were looking into machine learning and deep mural networks that were all the r at the time and we're trying to understand what can we do with uh guns how can we use them uh in our process to facilitate and augment what we do in the office so it was not even about J at the time or chpt M Journey so identifying things seeing the prospect of what you can do with them and reimagining how the industry can work that is part of our remit and that's what excites us and make us do all these incredible things that's where the Magic The Magic happens hopefully magic yes at least for us it is yeah sometimes it does feel like uh Magic when you hand these tools to designers so as rusc Clark said any sufficiently advanced technology is indistinguishable from Magic maybe talk about how the Ard team works with the project teams like for example does if I have a a problem do I come over and you know tap you on the shoulder and say I I have a problem with this script or something or are you kind of overviewing it the office from a kind of Broad View and seeing where you can implement we do both and I think that is where uh great things happen because you cannot we have been investing quite a lot in developing products and services within a that we deploy office-wide but of course you cannot deploy uh good enough products or Services if you do not really understand what your designers need and that's what we have been very very good at doing we are involved in projects on different stages for shorter kind of firefighting uh involvements or really long-term ones that may last for years and through that we can really understand where the PS of the designer lies what it is that they need what are the problems that they're facing and what are the best ways that we can solve them and what are the different ways that we can solve them as well so we do quite a lot of project support sitting with the designers understanding what they want and based on that we educate ourselves in what is needed and we build up our capabilities and in relation to the things that we we know if we put in their hands we'll make their job easier and their life better throughout that day I think one of the nice things about being on the tech side of an architect firm is you probably get to see all the products whereas some of the designers maybe just see this prodject so and then 6 months two years later they see this product so our office is really good at trying to create and an overview for everybody so people can have uh different um experiences and understand different projects Al together despite of where they're working on but I think yes absolutely what you say is right the fact that we are uh coming in as I would use the term support group but I don't really like the term but as a facilitator in many ways that is helping us see all different scales and all different stages of the project and also uh deal with a lot of different uh problems and issues and challenges uh which we find very uh interesting on a daily basis so you mentioned you've been at Fosters since 2006 you must have seen some I mean amazing development in technology over the years how have you seen technology grow but also play a role at Fosters over that period the interesting thing uh about Foster and partners that also had captured my imagination before joining the office was that te technology was at the heart of how the office operated that constant idea of Reinventing what architecture is and what it can do H is at the heart of what we do uh along with sustainability and a human Centric approach to design so that has always been the case and it's interesting because uh as an office we don't shy away from using new things trying new things and deploying new things and that is what makes us I think quite um interesting we have both the knowhow and the resource to do do that but also from the top down the the need and the support to keep on doing it so things obviously change as technology evolves but at our heart we have always been an office that has been really rooted in technological advancements you guys have created some amazing software and I've had a few little sneak peeks in uh shaped of fabrication when you showed a few of them you I think you've got Hydra Hermes and glaucon if I yeah if I pronounce them correctly how long does it take you to I I mean they're amazing and we have more than those just to bigly some some extra ones how long does it take to develop those and go from like the idea that we need this tool to implementation in the office it takes a long time until the point that you realize that you need to develop something because what happens is somebody comes with a question and a problem and you try to solve it and then you see two three 10 more people come with similar issues and then suddenly you realize you have something that is a replicatable issue that uh deserves a good solution at this point you start investigating what the solution might be and you start thinking what is out there what exists and how can I tap again in technology that can help me that does not necessarily exist within the industry today cyclopes which is our rate tracing software that is using GPU Computing and can run really really fast for example was born from this idea of how how can we do views very quickly and then how can we use them in real time so how can we develop an engine that can run in real time okay let's use GPU Computing as it's used for video and games and then realizing that okay now when we're doing views this is a rate racing exercise how can we extend that to environmental design so this is how these things develop on a technical level but at the root of these things is always the designer so what do they need what are the actual problems that we try to solve because I find that the biggest problem in the industry in people who do similar things within our industry is that we're always so enamored with what we're doing and how cool things are that we fail to see what is the actual problem that we try to solve and we're following on trying to just do things that are exciting so for us at the root of how we operate is um actually going out there speaking with designers and doing proper customer research in understanding what they need need and what are their roadblocks right now and what they would like without even telling them what we're trying to develop just by understanding whether the assumptions that we had on the software are correct and after we do this customer Discovery we start developing the software on top of that and we give them to trial and we're trying to understand what they right ux UI and all that but always with um with a mind towards how we can facilitate them and whether we're solving the right problem and I think this has been permeating the development of both our products and services for a few years now yeah you've built some amazing software will we ever see some of these outside of the walls of Fosters uh it's interesting what you're asking because uh many years ago this was this would not go uh this would not be a discussion the answer would be no but right now we are really investigating what the value of the software is in terms of opening it up to the world and always in conjunction to uh how this would potentially jeopardize our commercial IP and competitiveness in some cases there are software that we would like to keep in house but there are opportunities for software to be more exposed and these are conversations that we're now having with our executive board there's no answer yet so everything inside for now but we'll see how things evolve and develop as with everything in the office obviously I've got to talk to you about AI CU I I've uh yeah one we can't mention it at this era that we're in no I've listened to a few of the talks that you've you've brought it up um and maybe this will be a two-part question in one is how have you seen AI affects the design process so far and maybe um I know you already me mentioned the image generation side which is somewhat the kind of uh the new toy that everyone's playing with and then there's the kind of more interesting side of uh LGE language models and gen and all this kind of stuff how are you seeing AI play a role in in the design process I think AI is going to be um instrumental again on Reinventing How We Do design potential in the future but I also feel that right now in our industry there is an approach on a very superficial level and that can be problematic in the future because we need to really understand in depth what machine learning and AI can offer us throughout our process and how we can control the discussion and the situation so that this becomes a tool in our hands to enhance our creativity rather than replace it uh to that end we have been developing quite a few tools that have to do with surrogate modeling for example how do you train a model to predict analysis in real time rather than run the analytical process or we have been using a large language models to create our own search engines that search for our documents uh we have been using it for business purposes and even we have been looking into you know as you said image creation but not just random image creation but we have been developing applications that are incorporating with our within our design software like Rhino whereas a designer model something they can ask uh the AI agent to superimpose some sort of rendering on top of the model that actually exists rather than trying to imagine something uh I think it is important to understand that there will be a transform informational change in how Ai and machine learning uh are applied particularly with the creative Industries and I still I'm yet to see quite a lot of discourse around indepth issues uh on how we're using it in our process nobody is speaking really about copyright and IP which are significant issues and they're not necessarily something that is going to stop things from happening but we need to consider them as we move forward nobody is talking about what is the value that we as Architects are bringing on the table when Ai and machine learning are going to automate a lot of the things that potentially we're doing and how does this value need to be shifted there are a lot of very interesting things uh up for discussion right now and I hope there's more of that uh coming shortly yeah everyone that I speak to in practice is really about how can we utilize AI to enhance what we do I know there's a bit of a fear of people just typ in what they want they created an image and now I have a building of it but I think that's as you mentioned that's really where the future's going to be is this human and artificial intelligence combined creating something that we made it I've not seen before that's why I'm telling you everybody kind of talks about image creation or using chbt to get some answers right and I think this is just literally scratching the surface it's not going to it's going to be not even years even months where we're going we're already having models that can go go from text to 3D models from text to mesh that can automate processes that can create interfaces that can catch different apis and create workflows all these things are going to fundamentally change the way that we do our pipelines the way we work and it's going to be far beyond just Imaging and having an AI agent give us an answer to a question we need to really consider all these changes and see how we can we can be part of the change in order to see the change we want to bring forward in the industry yeah maybe on on that note not only on AI but like for a young architect listening to this and has dreams of one day joining your team uh are The Architects like that out there dreaming to join of course yeah I think definitely for myself when I was at uh uni I think Fosters had that kind of your specialist teams and I was always really Blown Away by the tech side of things and I didn't see that in every every firm so for sure there's Architects wanting to wanting to join your team I just ask the question because I like hearing it hearing the answer one what kind of qualities do you look for when you are hiring maybe young uh Architects have just graduated do they need to code and two with this rise of AI what would you advise for young Architects to get involved in this and not get kind of left behind so to speak so one of the major uh n inators of every hire is that they need to be good programmers they need to have at least good expertise in one language and we have certain language that we prefer over others but we're very open to that so this is something significant for us because that's the way that we communicate and we develop things with each other but I think in parallel to that what we have we value very very highly is potential you don't need to be the best programmer out there but you need to have the potential to know how to work well to understand things to bring ideas on the table and to like working on a team and even more importantly to see that the team is above all we're not really looking into having everybody showing themselves being lone wolves that are doing great stuff the power that applied around the enforcers I think has is that we're working as a team and if the team work great everybody's elevated and that's what we're looking uh into people that sense of potential that sense of com comaraderie and uh that kind of uh I don't know blink in the eye that you know they get it and they want to do something great yeah I really like your new uh Team shot where you're sitting on the ibeam in London I I I think it's important to I can say enough about the important of the importance of the team I mean my team in particular and the great talent that exists there we are all kind of coming to you speaking everybody who's speaking they have a team behind them we should not forget that we're not individuals who develop these things architecture is a group sport right and it's the people of the team that create the power for us to do great things so I can only say that my team is amazing every single one of them and that's why we are able to do great things and the same thing for foster and partners now going back to your AI uh discussion it's not only about a it's about every single technological advancement out there the trick is to actually understand in depth what it is that it's doing to not have a superficial view so that's what I would suggest don't try to just listen from your kind of uh reflective mirror of social media of what is important what is not delve inside yourself try to read as much as you can understand in depth what it is that any given technology is doing and what it has the potential of doing and only then you will really grasp the opportunities and the possibilities within the industry otherwise you're just saying again and again things that you you hear is like a neck CH chamber and that doesn't really help anybody so Final close as we look towards 2024 I feel we're in a very exciting time in the tech space there's AI going on game engines all this kind of stuff is there anything in particular you are personally excited about that you see on the horizon it is a very interesting question because I don't have a particular answer to that what excites me is the idea of change the idea that we can constantly reinvent the way that we do things that's what excites me and the tools are so many out there so many things that you don't even thought about a year ago and suddenly it becomes a thing that you absolutely need to see so I've never kind of uh picked up a player or two and said oh this is what you should look at this because I think if you are trying to play the medium uh it's a it's a bit of a risky profession um I want to effectively see how all these things that exist right now all the things that I mentioned that my group is already looking at how they're going to be able to differentiate our workflows change the way we approach design and how the entire profession is going to be is constantly evolving and it's going to be further involved through these Technologies so my ex my excitement is uh that sense of constant change and that's also what uh Norman Foster says all the time the only constant is change and that's what excites me I'm Adam Davis I'm a partner at Foster partners and Deputy head of the applied AR team so I think you uh you're also very unique like uh mafa you've you've been at Fosters since 2006 right so I think there's something really interesting how you've seen technology evolve over your time at Fosters particularly now we've got everything going on with AI and all these other kind of things but how have you seen technology evolve over your time at Fosters well uh I mean a major one major force is commoditization um so when I started um parametric design was still a niche Pursuit I think it's safe to say um we were creating bespoke parametric Design Systems uh Along Came Grasshopper And Dynamo uh and other systems besides um and you know we have had occasion to use all of them um um some of them in you know uh to considerable lengths and depths um and at one level it enables you know the the sort of the tide lifts all boats yeah it enables everyone not just within our team or within our practice but everyone in the industry everyone in the world um to do things with a speed and fluency uh that they previously wouldn't have had even we wouldn't have had partly because uh we were constantly having to build the tools that we were using yeah so it gave us a level of power uh an ability to address some very particular concerns but there was always this challenge um when you're both a tool maker and a tool user um of of having to wear both hats and do both things yeah simultaneously um so yes I think that commoditization uh and and it's parametric design is but one example um we've had some similar experience with um virtual reality uh extended reality um uh some of our former team members were uh working on virtual reality systems well over 20 years ago um with Hardware that was uh you know on the order of 50,000 pound a person for per seat per user um and you know we're now doing far more advanced things with headsets that cost a few hundred pounds um and so you know sometimes you have to wait quite a while for those you the parametric example was a smoother transition whereas um I think the XR one there was almost a trough or you know they talk about winters of Technology uh I think XR really did have quite a winter and and perhaps a Long Winter as these things go and then you know the green shoots all of a sudden just burst uh into into bloom or whatever whatever the metaphor you want to use um but it's just been this incredible um uh yeah just incredible acceleration of of technology in that space But you have a bit of a mix of areas of expertise you've kind of obviously got like computational and parametric backgrounds more recently am I correct you working also in the world of virtual Twins and this kind of space like what are the kind of areas of expertise you've had in Ard sure um I think the the way I would I would maybe try and the hook I would try and hang it on um is that I'm very interested in user Centric design people Centric design um so uh yes uh parametric design um very early on fairly early on in my time here um I began working with what we describe as quality of view um and developing um both tools but also standards um by which we might understand what it means to have a good view what it means to have a less good view what it means actually to have um um what we might call A A View a view annoyance or something something that you don't want to see um overlooking is is perhaps the most common example of that in most contexts people don't want to promote overlooking although interestingly as a practice we have had some situations recently with um designs where celebrities are going to be in attendance we're actually promoting you know a view of so and so or a glance of so and so becomes becomes an important activity um but nonetheless there's um understanding that in an objective manner uh and then there's also um opportunities to understand that in a subjective uh manner or trying maybe to understand not simply quality of view but qualities of view um so how would we describe it categorically um or maybe experientially rather than simply numerically um and the reason I say um the hook I might hang you know the the Arc of my career on it about being human Centric um my interest in digital twins uh although um it's a it's an area a discipline that's often you know very focused on um engineering on um you know mechanical and plant systems um on the kind of nuts and bolts of how a building is uh is operating um nonetheless I'm very interested in um how we as humans use the built environment um what we can measure about that yeah uh and how we can use those measurements then to do it better yeah um how we can uh design better for our public for our occupants uh for our inhabitants um and you know one of the interesting things about um user Centric design um the most important thing you can do is to ask your users um so actually uh you know do you like this do you like that um what would you like to see done better Etc um it's something that we're doing very much um in our software development work at the moment um and it's a practice that we're increasingly trying to apply to the built environment but the digital twin comes in because it's also important actually to uh observe people uh when they're not um thinking reflexively when they're not kind of hyperconscious or self-aware of what they're doing because uh you know it's it's well established that people tend to experience the world differently when they're in the process of reflecting and speaking about what they're doing or feeding back what they're doing as compared to just going about their business as normal so I think both of those um uh opportunities for learning about people for learning about what they do in the context to the built environment uh are very useful and the digital twin in particular um provides considerable opportunities for uh understanding human behavior uh in the aggregate um sort of unconscious uh human behavior um and given the nature of the physical environment that corresponds to people's behaviors um how can we understand those correlations um and and do useful design work based around those there's often an underrated part of the design process is actually getting feedback after you've built the thing it's almost never it doesn't really exist at this point uh before it does um but it is it is often difficult and it's not um it is it hasn't been a standard part of the way that we and the other disciplines that engage in constructing the built environment designing and constructing the built environment it's not part of standard practice yeah um so certainly it's it's it's by no means standard um there there there are precedents for it but they are um unfortunately rare and we're aiming to make them more common so maybe to take a step back and uh ask you for the audience that are watching this maybe there's some young students who are kind of you hear the term digital twin how would you define the term digital twin and are there different forms of digital twin for example here's a digital twin for maintenance whereas he's a digital twin of the world for whatever reason how would you define a digital twin so at its simplest a digital twin is a digital model of something in the real world uh and that thing might be an object or collection of objects it might also be a process uh or a system uh of of processes and objects um so that sounds fairly abstract at that level it is um more realistically and and more specific to the built environment um we're obviously interested in the data that is associated with land with buildings uh there's the the generic term assets which tends to cover them both um but it's it's the the things on which and in which we live and work and occupy um and data about how the those things uh come to be so uh construction you know digital twins have an important role in construction um how they operate so how they are used maintained um and I think uh often uh left out of that discussion or or perhaps a second class citizen for some but certainly not for me um is actually how they um are um decommissioned or disassembled or dismantled um and actually uh what happens to the after it's a building um for sustainability that's a huge concern um but I think uh you know there are other um potential aspects where where that information um becomes becomes relevant as well um and so you asked and that and that starts also this idea that um there's a digital twin that's relevant in construction um so you might think that actually all the constituent uh components assemblies uh Raw materials that form a building might themselves have a kind of atomic twin so your your windows your columns uh your concrete that's going to become slabs Etc um each of these things might have a digital being a digital Essence a digital record of its being that then um is joined fused together uh or or accounted for in some way as it coalesces as it gets assembled into the building that we're sitting in now for instance uh and then as those elements then eventually at some stage uh become disassembled hopefully reused potentially recycled um you could track the onward progress of those things uh you know ensuring that they're complying with relevant U sustainability guidelines or statutes or regulations yeah maybe a followup question to that is whilst there's no standard St software at this point for example digital twins do you see um an kind of off-the-shelf platform that you see as the kind of home for the future of digital twins for example unreal like more and more people are be building models unreal that could be used potentially as a digital twin for you know connecting it to iot and stuff like that or the Omniverse for example a lot of people experiment with on us is there a off-the-shelf tool at the moment that you see most use for digital twins well um I think a lot of what is marketed heavily in the digital twin space at the moment is focused on the visualization of twins which is uh important critical even um but it is not all that there is yeah um and we are very much interested in the whole spectrum of digital twin activities a lot of which uh has to do actually with with uh ontology so how uh what are the elements how do you define them and importantly how do you define their relationships within the twin um things like time series data your your iot data um and and then the visualization um and indeed not just visualization but um interaction with the digital aspects of the building um and that interaction is not only um you know potentially through a virtual walkth through um we're very much interested in augmented digital experiences of the physical space it's something we've been producing a lot of and working working through um in our uh development as a team yeah I think another referring to what I said earlier about uh commoditization um it is it is often helpful when um vendors in the market will make things that previous ly you had to do for yourself uh at considerable you know resource cost uh and and considerable development time when they make those things simpler and easier uh so I do think uh and you know some of your examples there are already uh some systems that make some of those things easier and indeed and indeed we are using them uh people often speak of software as an ecosystem and I think an ecosystem thrives when it has uh many participants um that have a variety of different interactions with one another uh so we would very much like to see across all the software that we use but in digital twins certainly um this ability for different systems to interact in useful ways uh rather than looking for any one system to provide everything that we might need from that yeah great one question I was really interested to ask you uh like Maas you kind of sailed through the parametric computational era and saw that evolve uh at a very unique practice that was at The Cutting Edge of this now we're here in this AI era which seems to have a lot of parallels uh but they're also two very different beasts do you see any uh parallels in the parametric computational era and the era we're in let's say the the AI era uh that we're in right now I see some parallels I think um both of them make digital manipulation of the models that we create uh of our understanding of the built environment both of them make that in some way more immediate more um accessible to designers um there is a a way in which I guess you might say they they act as a sort of force multiplier yeah um so so so they they make uh they offer the potential to make individual designers in some ways more productive okay um I think uh critically AI um particularly with its interfaces that are oriented around natural language at the moment but it but I think I'm sure increasingly we will see systems that are able to for instance interact with uh a user in almost a sort of napkin sketching mode I mean I know there are some systems that already do that but they're they're not really um commercial grade production ready at the moment but they will come um and as compared to um parametric design it's potentially going in the other direction I think parametric design um helped some users some designers to think more logically more abstractly about what they were trying to do um more on the meta level of what their design was trying to achieve and how they could structure that as a system of rules um that would operate on data parameters uh to bring about intended ends I think uh AI is offering systems that try to infer what a user's wishes might be right I do think to some extent I mean the the the title the sort of job role prompt engineer suggests that actually what people are doing is thinking analytically about language and how the effects that that language produces uh will change output will result in desired changes to Output so at that level there's there's still that structured thinking um certainly the engagement with that structured thinking as computer code I think is is is being pushed more into the background um I think there are good aspects to that yeah um but I also wonder if some of the structured thinking that goes into making a computer program when it is put through the channel of uh human natural language yeah or natural modes of representation like drawing and sketching whether some of that Precision of thinking may be delegated to the computer rather than being being actually taken on increasingly on the part of the designer yeah very interesting yeah no I think that there is a real nice point that you're touching on there is is like some of these tools give us the quick access to these things but they take away sometimes that deeper thinking process yeah I'm Ked I'm part of the applied r&t team I'm a partner and my interest uh combines uh some complex geometry so I work on like a lot of uh large projects in the past and uh also robotics which is like uh a personal interest of mine I've been doing research in in it for for a long time and performance driven design so some of the tools that I develop and I use daily are related to how we can kind of find the best option let's say and how we can grade our designs in the office sounds like you're covering some interesting areas performance driven design robotics we'll get into some of those but maybe talk to us a little bit about performance- driven design maybe from some young Architects that are not sure exactly what that entails what kind of things do you do with performance uh driven design in terms of on projects how does that actually work so for us it's always a question of how you can Rate Your Design option so and and it's like a a long discussions as we have the teams usually is like we we go through Millions options for for each project uh and the idea is you try to find criteria that you can grade your option in and this will be become like grading the performance of each option uh and basically you try to find this comes through either analysis uh that we run stuff like view analysis uh performance analysis in terms of uh daylight uh energy and so on or even stuff like uh financial data uh like whatever whatever the input is for each option we try to find the best way to uh graded let's say maybe let's talk about for performance in uh performance D design what kind of tools do you use for that like is that I know you guys will probably be utilizing grasshopper and coding but you guys have also bu in-house tools stalia if I pronounced it correctly uh maybe talk about some of these tools and how you use them so uh as you know we build our own tools in the office and this is something I've been doing for a long time a lot of them are long-term developments so something like Thalia we have been doing I think for for five years now uh and it all starts with an idea and on a project and then it takes its own life and then we start having like a a proper development uh uh uh life for it that we can go through different and so on sothia basically tries to combine uh modeling and Analysis in the same interface so it can imagine somewhere that you can have direct feedback uh for whatever option you're modeling very quickly and this thus you try to get like uh um uh through your ideas much faster let's say uh and uh this all started by um the development of some of the tools that we're doing like Cyclops and so on some of the very fast analysis and combination of like gaming engines technology that meant that we can have analysis that needed like in days or a cycle that needs like two or three days to go through a specialist now you'll be able to just have that directly in your uh interface and this becomes like the kind uh uh what's the word uh let's say it becomes a power of this Tool uh uh where you can discuss your options with uh your colleagues for example or discuss your options with the client directly while you seeing while you can see all the performance criteria for the options that you're presenting let's say you've essentially got a personal super assistant uh helping you whilst you're designing in a way yeah exactly it's I mean this is a good Focus for us we we need to make sure that we're like producing the best option possible for the uh the project that we're working on and this also becomes a conversation because there's it's never one uh there's never one option that's like covers all the criteria right it's always like a balance between three or four and then you need to make sure that you presenting all of these uh in like let's say Unified interface that you can actually try to look at the options and see which one performs better for you so I see you're also involved in the robotic side uh also specifically spot the dog uh I've seen you in a few videos with uh the famous spot the dog maybe talk about uh why are archtics experimenting where with tools like or robotics like spot the dog what are you guys using it for and where do you see it's advantageous for otics to utilize yeah I mean robotics have been an interest with mine for for a long time and uh it's great to see how some of the technologies have evolved let's say some of them is the way robots can really navigate space this is like the most important part that I that we saw in uh in spot and its development for example uh so it have been a problem for a long time of how uh any robot can navigate the space that the humans can use and how it can actually interact with humans as well in the same space uh and when spoton came out with with its abilities to like uh navigate terrains go upstairs and all this kind of stuff we saw directly uh how it can be beneficial in in in in architecture for sure and what happened this all started with a conversation was a GU was Dynamics uh in a conference and then we got to the idea of let's try to test how we can use spots for trying to monitor uh construction sites and we started to build the idea of how we can use it around the construction site daily basically you get a history of how uh the construction site is progressing through time and then you can do all your checks on your uh Bim model later on make sure that uh everything is bur corre correctly and if there is any problem is getting mitigated early on so the whole idea for us was trying to build this workflow that like goes from scanning uh uh to checking the data through against the B model to getting a report in the end of it and seeing how this all can be automated so instead of having it you know scanning the site uh once a month whatever someone go do it manually we something is done automatically every day U and then you just see the timeline for the changes uh through the whole life of the construction site no it's it's a very surreal thing to see it walking around I'm still looking around to see if I might see it walk down the uh I mean you can imagine the first time quite noisy right it's it's a it's a machine it is a machine and it's very weird because when people see it and we used to walk it around as well here and at the time it was still AA so it was like there was I think 50 of them and we had like one uh and you can see actually lots of people coming to P it all the time taking selfies with it even the construction site so it wasn't like intimidating by any mean think the way it works and the way it's modeled around like all the Kines that we like and so on made it like a very approachable kind of machine then actually not that scary let's say even with sise and all the stuff that's around it um I think still works around humans quite well yeah uh and and this was also a surprise for us because we thought it would be uh what's the question how's the people on the construction sign going to deal with it are they going see it as a threat is it going to be uh uh uh uh like for them something that we were going to try to avoid or so but on the contrary it was all uh they all kind of welcomed it to the environment let's say yeah yeah I think that's that's super interesting and we're in an era where I think uh obviously robotics seems to be on the rise with you know powered potentially by AI do you think we'll see more robotics in the construction industry particularly with the rise of human robots where you have Tesla Boston Dynamics creating more you know human like robots do you think we'll see that soon or in the not too distance in construction sites no no I think it will it's more and more you can see even I think HP now have a robot that you can that goes around and can draw like uh uh lines on the site and so on so you will see a lot of Comm commercialization of a lot of these stots uh uh very soon I mean what everyone is trying to do right now is trying to find the best form factor for robot that can fit everywhere the humans does so that's why humanoid throt would work perfectly because you don't have to redesign your process around it it will actually like just get integrated in whatever process you have already uh again still the whole problem with navigation and uh computer vision is still like I would say the biggest hurdle but with all the stuff we Advanced we see see in machine learning and so on that won't be very long so I would say we're very near seeing this kind of robots around us constru construction sites at home um definitely like it's kind of will all come together machine learning bit and a uh kind of robotic advancement as well yeah uh and I'm assuming this will happen like very near like the next three four years for sure we seem to be at that inflection point of these two worlds marrying and taking off it's going to be exciting future yeah yeah you can see like autonomous vehicles all this kind of stuff just uh uh you can see when you start having all these computers like inside a a card that can be controlled by uh an algorithm and then you all the stuff start coming uh together let's say yeah hi everyone my name is Marius um I'm a partner at the applied R&D team uh where I lead uh the interoperability Endeavors of the team and design collaboration you want to call it um but apart from that I would would say I'm also like uh jack of all trades I've been uh using design computation for many years and yeah I mean this practice and our team in specific is uh the best the safe platform to start experimenting more with technology within our industry I'll throw you a tiny bit of a curveball question to begin with but I liked on your on your Instagram you have ex Architects right and but you talk to us a little bit about your background like you you trained as an architect now you're more a a coder developer side talk to us about your background and do you feel that now you've bridged you've crossed the bridge of uh from an architect to like a developer so to speak or so yeah I mean I was originally trained as an architect and then started you know being interested into the complex geometry firstly so that's how uh I gone into the world of let's say scripting first and very uh all days uh of visual basing scripting I suppose uh but the reason that why I have like X architect on my title is that I don't think I design in the same way uh as I did before as I did during my studies and as you know traditional archtics are doing um I do feel those that we design as well you know being integrated into the project delivery process or um within an architectural practice anything that you do even as a as code AS writing code at the end of the day has an impact on on a building right so in that sense I still feel that I'm designing but not in the Trad traditional uh uh way yeah uh I I would say that lately yes inclining more to software development but again this softare development is very specific to the to the industry obviously you might have heard before that we're using all sorts of Technologies uh from web Technologies to film uh or game engines uh so we try to integrated within our tools uh so it's it's a constant learning uh process uh and that's the the thing I I enjoy more and computational design is one of those ones where it is designed like like you said even though you're coding or running lines of code or even in grasshopper it's still having a big effect on the design so no absolutely yeah it gives you the opportunity to explore the design space firstly uh what you know a simple parametric model can give you thousands of options that you can then if you choose you can analyze and assess and hence inform your decision making right um it does have an impact and uh I do believe that most of the current let's say uh young Architects that join the practice and especially because the academ Academia is also leaning towards let's say this uh build is maybe a little bit bit hot at the moment or it is hot for 10 years now um they come prepared and they have an arsenal of tools that they can use right U so I would expect that every single member that joins any design team in the practice now has uh a basic knowledge of how to use visual programming uh interfaces like grass opener for instance and set up a a very simple parametric model or in a way like take a parametric model from someone else maybe our team yeah and start you know adding to that or altering it or tailoring it to its own needs so it's very important that you know people come in with this skill set yeah because it also allows us as a team to um focus more on what comes next yeah in terms of tools in terms of processes in terms of workflows so maybe talk about how do you guys how do you work with the design teams and projects is it like a as simple as someone coming up over and tapping you on the shoulder and it's like got a messy grasshopper script all the way from creating much more advanced workflows how do you work with the teams particularly with computational design yeah well there are a couple of scenarios in this in this case but it's basically what you described like people come to us yeah with a problem usually and uh they come to us to provide a streamline solution to to this uh problem maybe they know that we already have a solution or they're engaging with us to create a solution for this so that can happen like in the very first stages of design or maybe on competition like stages or it can even happen like later like when you creating documentation for uh for uh construction drawings for instance uh the level of uh involvement that we have depends on many parameters uh and the main one is I would say the complexity of the of the problem when usually when we involved in early stages we tend to continue uh because we set up workflows for the team to work more effectively um whatever this work CLS might be simple parametric models paired with analysis or optimization uh uh studies running in the background as each iteration of the design moves forward uh but I would say that we would definitely like to be involved quite early yeah as we can provide this um let's say uh feedback that is paired with data to come with it yeah uh so we can provide let's say like some sort of um assurance that a design decision that's been made is supported so usually becomes it starts uh either with a form that they want to achieve if we were to strictly talk about uh architectural geometry I would say and how it's better to achieve it how we can integrate the fabrication constraints let's say how we can optimize it to be cost efficient um it's rare that we receive let's say uh workflows uh from our design teams because as you may know very well it's easier to build something from scratch rather than start hopping upon someone else's work yeah and that's the beauty of the systems right um there is a million different ways of doing the same thing right and each one of us has uh and within the team even we have our own you know uh workflows we start from something else but we always result to to uh something that is similar I suppose so yeah I would uh I would be not against receiving U parametric definitions from other people but you know a little bit too focused and I would actually oh yeah you know this is not how I want it to be I'm going to change it a little bit ah yeah this is like is um controlled by some data that I don't like it's like peering into someone's thinking process as well yeah exactly no it's it's as I said there are million ways to do the same the same thing it's uh yeah it's a deterministic process in the end but the in between is um something that is a bit hazy yeah I suppose so you mentioned interoperability maybe we'll talk a little bit more about that I know you're involved in the development of of Hermes your inhouse inter inter oper IL tool talk a little about what that is are we talking inter interoperability between Rhino and Revit for example yeah what is the goal of it so yeah I mean Hermes if I were to um give it let's say one line let's say description this is a more of a messenging service mhm uh it's cloudbased yeah uh it does interoperability at the same time because you can tap upon like various proprietary uh applications you can create clients for those and you can send and receive data so what Hermes does is basically connecting our designers their machines across like the locations of our practice um real time and and On Demand let's say so it it facilitates this uh exchange of of design data M and uh it is very important that these data are like very lightweight they are also secured the version as well so you can actually go back to something you've done like let's say a couple of days ago and you have a version of this that actually works still or um or you can audit let's say this data um so I would say like Hermes is more the facility Ator what is important in this case is also the glue or the the the language of data that are been uh exchanged uh with this system and specifically for us as a team they're very important because they connect our other products together so it's one of our other products I'm pretty sure like Martha talked about some of those uh they all speak to each other using this common language okay so it's much more broad it's not like interop between A and B it's kind of agnostic of what corre correct it's I mean there are very there are many tools out there yeah that do the interoperability side very well and we as a team we always keen on uh picking up tools that are already there and they doing their job very well like for instance uh R inside yeah we we been using it a lot uh how do you integrate these tools in your workflows mhm mhm and how you integrate these workflows with the design teams and how you have this glue in between them that actually shares data yeah is really hermes's job so you we we're not think in a traditional way of oh yeah I have a very complex geometry this double curve right how does this become a Revit object right yeah yeah this for me this is a easy part what happens in between and how you can access this data is the the that's not fun part I would say because it's more like a backand infrastructure because no one sees it yeah um and Cloud infrastructure which also very tedious uh but uh yeah it's it's very important and potentially it will be more important when we have a consistent let's say formatting for data in our industry and we can train models yeah and we can utilize these lightweight formats like other indust have done so far so we've been asking everyone that we go to from an incredibly broad perspective of the software that you use so what you use design is it Rhino or or Revit for for Bim but also specifically at Fosters you have this incredible Suite of tools so maybe give us an overview broadly of the tools that you use and also inhouse the tools that you have to supplement them um yeah I mean obviously cat packages if we were to Cluster them or to like put them all in the same category some of them are good some things others are good at other things and obviously rabbit or B packages are are very good documentation right design documentation I means it's a no-brainer you cannot really generate sheets uh construction documents manually anymore it's it's pointless so yeah think the the office is using uh beam a lot there's a very big uh Beam Team called BDS building Design Systems um but also like most of the designers are using uh Rhino for uh for sketching modeling on all levels I would say yeah then it comes the discussion how this talk to each other how do we bring you know complex geometric forms into something that can be documented annotated tagged Etc so this is where the trouble happens yeah and um but yeah nothing we can all solve I suppose um uh apart from that I I haven't really used any visualization software myself I know the teams uh every every workstation of the technical stuff design stuff has some uh let's say rendering engines uh tscape might be one learn more it doesn't really it depends on the person what they're comfortable with using uh I wouldn't say there is like the recommended uh software if you can make something look beautiful in one piece of software uh there going to be the opportunity for you to have the software on your on your workstation for instance and what we as a team are doing is trying to enhance the design experience really with our tools so simply by uh bridging cat packages together with Hermes or providing an intuitive meling interface with direct analysis with Alia or running analysis as you design with Cyclops in Rhino or um running optimization in the back the background with Hydra so our tools are meant to to come in in the design process and enhance the capabilities of the designer and ultimately I would say like save time uh allow the designers to focus more on what's needed in terms of the creative side yeah of the process and you know stop annotating let's say Flor plants for instance right I mean I mean I think most of the members let team have the same uh belief as I do like if you do something more than three times you need to write something to automate it I do have my own automations as I work for instance maybe I'm too lazy or something uh but that's what we're trying to uh provide for the whole office as well yeah if something can be automated we'll make sure that we pinpointed and we're going to we're going to provide the solution to that again we're trying to make their the design experience as as smooth as possible yeah and as uninterrupted undisrupted maybe as possible yeah with our tools I think it's that nice it's definitely a reoccurring theme of freeing up designers to spend more time thinking and less time producing or repeating things yeah absolutely I mean there are many talented people in our office and in other practices and it's it's waste for this talent to be uh copying a thousand trees around ex why would you copy that there is a process of doing it automatically yeah yeah and obviously you know like new systems hopefully large langage models maybe going to assist this even better if it can be integrated in this workflows the last closing question is I know you teach as well at uh the butler right um what kind of advice would you give for a young aspiring architector may be interested in getting into computational design I would say that you know computation it's it's not I mean it has various let's say um levels yeah how involved you want to get um for me uh any student nowadays should have a basic understanding of uh at least parameter modeling yeah uh I think it's a great asset that uh you know someone that's going to go in a couple of months in a couple of years out in the in the market to to look for a job uh can have right it's definitely appreciated and sometimes it's required yeah uh no one at least but in my knowledge has gone into an interview and then someone told oh yeah you you're doing parametric models oh yeah we don't need that yeah that's like not going to happen right um so it's it's essential for me to have a basic understanding then how much in depth each person will go into that it's up to them not everyone needs to be um a computational designer or um design computation Specialist or how's it called uh specialist in any way right yeah there might be some people that set up workflows and you must know how to use them on how to make them your own yeah you don't need to set up everything from scratch [Music] so I would say as you learn other tools it would be nice to start like looking into these tools and this way of thinking because it is a different way of thinking it's not as intuitive as you know Strokes on on a drawing board or like pulling uh polygon meshes in other modeling software for instance um so it it is a more steep learning curve because you have to change your way the way that you design in a way the way that you think you have to think of the result first and then how am I getting there yeah and plan this uh but it's definitely very exciting and for me again the most exciting part is that you can reach the same conclusion in a completely different way than others um it's a constant learning process and that's why I also like enjoy what I'm doing every day I'm learning something new every time uh our designers come with a different problem we haven't seen before so it's very engaging it's very engaging I hope more people go into computation uh more people save time uh for their uh colleagues I I would love to save time to my colleagues that are great designers right yeah yeah yeah um so yeah my advice is just start picking up the tool know how to use it in a basic level and then decide what you want to do further down the line yeah but definitely pick it up definitely get that computational grasshopper base yeah base is essential it's it's a great tool to have even if you just use it to create like folders in your machine you can do it really yeah you can really do the simple thing it also saves you time yeah all right guys thank you for joining us on this architect tour and a huge thank you to The Fosters Ard team mafa Marius Ked Adam you've been incredible I'm totally Blown Away uh I think what you have here is incredible uh stay tuned for our next episode and if you enjoy this video give us a like And subscribe because it super helps the channel otherwise we'll see you in the next tour

Channel: ArchiTech Network

Views: 20,284

Keywords: Architecture, Architect, Technology, ArchiTech, ATN, computation, bim, visualization, graphic design, immersive, augmented reality, virtual reality, game engines, parametric architecture, grasshopper3d, parametric, metaverse, Space Architecture, 3D Printing, AI, AIarchitecture, Artificial Intelligence, artificial intelligence, big, bjarke ingels, bjarke ingels group, big architects, office tour

Id: p-oF3aIG_OA

Length: 83min 26sec (5006 seconds)

Published: Fri Jun 28 2024