Recreating the Original Vision for Habitat 67 with Unreal Engine

JARON LUBIN: Good morning, good afternoon. My name is Jaron Lubin. I'm here with Carlos Cristerna. And we're going to talk a little bit about our presentation for AU 2023, excited to be here virtually.

And our topic is recreating the original vision for Habitat 67 with Unreal Engine. So my firm, Safdie Architects, has been working with Carlos Cristerna, who's co-presenting with me today on this wonderful project, sometimes referred to as project Hillside.

Carlos is from Epic Games and he'll talk in a little bit and share some of the work that we've been doing collaboratively over the last, I think, Carlos, it's been probably two years at this point, right?

CARLOS CRISTERNA: Correct.

JARON LUBIN: So I'll just jump in. So, Safdie Architects, we were founded in the 1960s, actually. Many people may be very familiar with the project, Habitat 67, which is in Montreal. This is a construction photo from the '60s, as soon as Habitat opened. And Habitat was built as an experimental housing project as part of the Expo 67 in Montreal, a World's Fair.

And it was a radical rethinking of how you could live in the city, in a dense urban environment, and also a radical rethinking of how you build tall buildings, or buildings for that matter, using prefabrication modules, in this case, volumetric construction that were built in a factory, brought to the site, and then hoisted into position, and dropped into place.

And so this was, for the time, extremely experimental and, ultimately, became this fairly iconic project, that, if you're from Montreal, are very familiar with.

These are some recent, or fairly recent photos from above. And this was Moshe at his desk at McGill University in 1963 or 1962, working on his college thesis.

And I should say that this radical idea for this wonderful project which many architecture students and many people are familiar with in history, this was actually a college thesis project which started as a series of doodles and experimental models. And while the project, as I just shared it on the screen as built, many people are familiar with and have even visited.

Many people may not know that the original idea for Habitat was much bigger. And it shared similar principles, and guidelines, and aspirations of a three-dimensional city built out of this kind of volumetric construction. And there was a series of these-- let me just-- there was a series of systems that were thought of, and Moshe called them systems A, B, and C, in a kind of shorthand.

The bigger vision, aside from rethinking how you might build, was how you might live. In other words, the thesis for the project became, or was generated, for everyone a garden. So everyone had these big outdoor terraces and access to daylight. And it was like bringing some of the characteristics of the village, Hilltown, or the things that you would maybe celebrate if you were to live in the suburbs, like having a big backyard, into the city.

So you have all these amenities of access to air, and outdoor gardens, and these terraces, and daylight, if, even, you were to live in a dense urban environment. Here's some of these wonderful sketches that he created as part of his college thesis, trying to imagine what this could be like.

And these embryonic sketches started to form around these different systems, whether they were a bunch of sticks with these kind of modules fitting in, whether they started to organize as what we call a series of membranes and surfaces of these stacked volumes. And at the same time, thinking about the conceptual rethinking but also solving this for construction. How could it be very constructable? How could it be very deliverable-- and quickly-- using prefabrication technology?

And there's these ideas that were emerging that if you were to build these housing prototypes out of these modules, out of these rectilinear extrusions stacked on top of each other, how would you connect them? So you have to rethink everything. So you don't have corridors and double-loaded corridors that are enclosed with no daylight. You have these open concept corridors, which are more like streets in the sky.

And, OK, you need rain protection. So there's these enclosures over them. And you can imagine that these are shared spaces. So they're almost like an extension of the city streets and the sidewalks below. And how do you stack these things geometrically so that each volume and these streets knit together so that your neighbor's roof becomes your private terrace? And it's all this elaborate kind of three-dimensional game.

And, of course, doing this in the mid '60s, this is done with physical models and with LEGOs. And think about it. There's no computers. There's no Rhino and three-dimensional modeling. And even at the time, LEGO was new.

So Moshe hears with his team, there's this new toy. And they start playing around with this module to start playing around with, how do you how do you solve this geometrically? Large physical models were created in an office that was created also. Moshe immediately formed a small practice with friends and professionals that were brought in to solve for this idea that was generating a lot of buzz and that was part of this commission to build for the World Expo.

I should say that Habit, the difference between what was built, which is purely residential, and like I said at the beginning of this introduction, what was originally planned. What was originally planned here was a mixed-use development, actually.

So what you see on the screen here was this finished model, this conceptual model of what was considered Habitat.

It has a hotel. There was residential units. There's schools. There's retail, all floating underneath this membrane structure and all connected to a series of pedestrianized streets, both at the ground, in the air, separating people and cars. And with all of the units actually having celebrated views to the water, this really beautiful river site, and for everyone a garden.

And so, Carlos, now, via way of introduction, I've been working with Carlos in a certain capacity for over 15 years now. And he brought to our attention this idea that maybe we could actually, even as this original larger-scale idea, which was not built. What was built was actually on the right hand side of the screen, a variation on the residential units that you can see on the right.

But Carlos had this thought that maybe we could actually, through new technology and through some of the things he was playing around with, bring the original vision and the original dream to life virtually, in three dimensions, using new software and to show the market, to show the hotel, to show the office, the art museum, the school, the entire landscape framework.

And so we started to work together iteratively, back and forth to construct a digital model of the original based on archive drawings, archive information. And there was a lot of holes and kind of things to solve. So, actually, the design team, between our team here at Safdie and Carlos's team, had a lot of fun bringing this dream back to reality, albeit in the digital space.

And so we now, which is really wonderful, had available after, maybe, a half a year of work, and I'll let Carlos talk about the whole process and the whole collaboration. We now have this digital recreation using the original drawings of what could have been.

And so we were all very excited about the potential of the literal model on the left, of this radical idea of Habitat, and what was starting to emerge with Carlos' team, what he was calling Project Hillside on the right. And with that, I'll share it with Carlos. And he can take you through what happened next.

CARLOS CRISTERNA: Hi, thanks, Jaron. This is project Hillside right now that I have open. I'm showing in Unreal Engine. And like Jaron said, we've been working together for a while, doing architectural visualization for Safdie Architects.

And then when I joined Epic, and with the advent of Unreal Engine 5, and having done so many projects the traditional way, if you will, on architectural visualization, the idea came to mind to bring a project to life using new technology, new real time rendering technology that could serve many purposes. That could have some historic weight to it and some educational weight to it, as well.

Originally, we just wanted to do, here at Epic, an architectural-visualization demo that we could just share with everyone so people could learn how to do architectural visualization in real time with our tools. But we figured that needed a little bit more soul, that we needed to get a little bit more weight to it.

So having had worked with Safdie Architects for a long time, and having been familiar with their portfolio of built and unbuilt material, Jaron, and his partners, and I, we sat down one day.

And we did a brief sales pitch in which we said, hey, how about we bring one of those unbuilt projects to life? And we just give it away. We just give it away to people to study, to students to study, to people to learn how to do architectural visualization and all of that.

We had a couple of candidates to build or to develop. But the obvious choice was Habitat 67. That was the project that gave Safdie Architects and Moshe Safdie, personally, notoriety back in the '60s. And it's a project and principles that Jaron and the design team at Safdie continue to chase.

So I figured, hey, that has a historic component that has the educational component to it. And let's just go for it, see what it looks like. I've seen those physical models in person that Jaron showed in the slides. And it's fascinating to see them and to see that they're well preserved. And they're beautiful and well done.

And you're like, man, I wish we could see this building. I wish we could see it in real life. So this, literally, is why we did it. And what I want to do now is I just want to show you guys this project a little bit more closely. And maybe Jaron and I can go a little back and forth with some of the things that we wanted to do and the challenges of doing this.

So what you're seeing here is-- I'm just going to go through a cycle, a couple of views here so we can see. The real time version of this model, this is the model that you can download from our marketplace and in a version of Unreal Engine that is also a downloadable version of the marketplace.

And we wanted to test, us at Epic wanted to test the limits of what could possibly be done with it. So we chose a mixed-use development that obviously was a very, very big project. You see the islands in Montreal over there. There's plenty of detail here on this model.

And the process was quite interesting because we had to come up with a new process. We worked with an architectural visualization studio called Neoscape. Some of you might know it.

And for over 20 years, we've been creating linear content for [INAUDIBLE] a certain way, with offline rendering, and things like that. And working with architecture and design process can be challenging. We wanted to see if using new tools, new technology, we could make this process, the creative process, a bit more flexible for all the parties involved.

So when Jaron's team was working on, really, designing the building because all they have at McGill was the archives of the two-dimensional drawings done by hand, physical models and it was a concept. It was not really solved, especially at the ground plane. We had to come up with all of the landscaping and all of those street levels and connections.

And how are the units going to stack up when you have the physical model and you see it? It's, at the end of the day, a massive model. But you have to come up with the rhythm, the windows, the sizes. How big are they? How many of them? The gardens, how are they going to be [INAUDIBLE]? And what kind of vegetation are we going to have, et cetera, et cetera?

And all of these different pieces are, obviously, it being a modular design, one may think, well, it's just repeating the same unit all over and over and over. And while it may look like that, it's not that. There are several different variations of that unit that make the building.

So let me show you quickly what I mean by we. The design team really had to work on all of these transitions, and landscaping open spaces. It was very fun to see these for the first time.

As we were working on the process and we were reviewing these with the design team, particularly with Moshe Safdie, all of us were very happy to see-- because all of these ideas and concepts we worked on in the past, I heard them mention all of these things about openness and a garden for everyone, and the membranes, really, trying to enhance the quality of life with design through design and how open these spaces really feel, and not oppressive, even though we have a large scale mixed-use development with how many units? Is it 1,200 units, more or less?

JARON LUBIN: Yeah, consider it around 1,200 to 1,300 resi units.

[INTERPOSING VOICES]

CARLOS CRISTERNA: --go back and forth.

JARON LUBIN: It's interesting, Carlos. It's like it's a common brief that many architects are facing these days, which is approximately that scale on small urban sites. So it's really kind of timely.

CARLOS CRISTERNA: Yeah, and so that was part of the reason why, as an architecture student, I figured, hey, this seems like a good project to take a look at and really analyze the merits of the ideas and study them.

We were thinking, there's so many people that A, didn't know that there was a much larger idea, that all of these principles that Moshe continues to chase, and your practice continues to chase in implementing projects all over the world now, are valid. And all of contemporary architects are designing like that because the world is getting a little bit more difficult to live in because we know that the American suburb or the urban sprawl is not sustainable. But we keep doing that.

So it, to me, that was like the architect nerd inside me wanted to get some of that put out there digitally, so that every single architecture student had a chance to take a look at it, and study it, and take what they will from it.

And, of course, there was the idea of, process-wise, coming up with the usual deliverables of architectural visualization. We're going to make an animation. And we want to make a bunch of still images. And we want to make an interactive.

There was an interactive that we made that you can download at anytime from the Unreal Engine marketplace and the Hillside website, that in this interactive experience, you literally walk, get a personal tour of the project explained to you by Moshe Safdie.

There's, for instance, we had a very nice conversation at his office when we were going over the project itself. And then we would ask a question. So, Moshe, we're standing right here, right next to the hotel. And we're looking down the spine, the main axis of this project.

Tell me, what is it? What are we looking at? And why are things the way they are? So that gives you an idea of this personal tour that Moshe gives you with his own words. And to me, that was a proof of concept of the new architectural textbook. You want to have these things documented and live forever so that people can always go back to them, listen to them, and understand where this comes from.

People are quick to judge architecture by looks. We're all quick to judge architecture, or books, right? They don't say, judge the book by the cover. But when you hear an architect, a designer like Jaron and Moshe, and they tell you, and you're looking at it and they're telling you, look, we want everyone to live like this, everyone, every single unit has a beautiful garden, every single one of them.

And there you're thinking about quality of life. You're thinking about light. And you're thinking about people, not just having a little balcony. So you understand architecture differently when you start doing these things. And that really was the gist of why we wanted to do this. And think it was very much emotionally loaded, wouldn't you say, Jaron?

JARON LUBIN: Totally, the thing that got us is when Carlos came by one day. And I should say, by the way, that the team that worked on this with Carlos at Epic and Neoscape, they went beyond, in our opinion, what's typical of an architectural visualization by the way. There was a lot of atmosphere, and emotion in much of this.

And what caught us was actually when you came by and you started to show us the kind of side-by-side imagery that you had assembled of like taking original archive photos, and drawings, and illustrations, and putting them side by side with the kind of contemporary Unreal Engine-generated view of the same angle.

And that was it. That was game over for us because what we realized was the potential of how this could be an educational tool. How this could be amazing for architectural visualization and people in that industry. And, also, how it could be useful for us in projects where, aside from this effort, that we're in this situation-- designers today-- where clients are demanding this level of sophistication to talk about our work and to show our work.

And it's becoming increasingly difficult to do it because we don't have as much time to deliver these things. So to see how beautiful these images could be created in real time, this seemed to be, maybe, a glimpse into new workflows. Also, in terms of we're also thinking like, oh, all right. How can we use this in our process for many products?

So we'd imagine that there's something here that could be super helpful for students in school studying architecture on their own projects, not just to learn about Hillside and Habitat, but also we were thinking, could be helpful for their own design work and professionally for our design work, to help us out with the many things that we're into.

Obviously, we were very excited to work in this way with Carlos and the team. And this is kind of like the effects. I mean, you guys, Carlos were really pushing the limits in testing. When you were zooming in on that water, you're showing off, right? That's something that's super-- that goes beyond the typical arc [INAUDIBLE] situation, right? I think that's what you were trying to push.

CARLOS CRISTERNA: Yeah, I think we're trying to-- the effort was we were shooting high. We're testing our technology. We're making sure that we actually can produce something that is what you said. Every single developer and every single client of yours wants more and more and more because that's the way for better or for worse. Architectural visualization has evolved. It's a very powerful sales tool. And people use it all the time now.

And so that was the one side of it. And then there's the other is the educational side of it. There's the testing all the technology, like I said. There's does it work? Does it not work? Can people do it easily?

And that's one of the things that we wanted to do here, as well. I'll briefly show some of the organization that we made for this is quite clean and quite neat.

And usually on Real Engine demos, are very much technical. And this one, while it is somehow technical, we did the very best we could to use the tools as they are there. There's nothing special really done to it.

We're just following best practices of nice organization of clean models of nothing too fancy, so that every single artist out there and every single architect out there can download this, follow the documentation that we put together for you guys, and really learn exactly how we did this. I'll give you a quick example before I move on to another part of the project that I think that we are overlooking, which is Habitat 67.

In this file, we have what we call Project Hillside, which is what we're looking at, the original vision. But we also laser scanned and used photogrammetry to preserve the build structure digitally and for people to have a chance to explore it on their own when you don't have access to going to Montreal and see it and things like that. But here's a very quick example of what I mean.

So we have nice and neat organization with the environment lighting over there, special effects on their folder. Everything is named properly.

These are going to be your three folders that are the learning ones, if you will. Here in maps, basically, you have the map that shows the exterior, which is Project Hillside, the map that shows the interior, which is Habitat 67. And then a couple of other maps that have motion graphics and things like that.

And then you're going to find folders clearly named here in SA for Safdie Architects. It's basically all of the different components that make this scene. And you can explore them one by one. You can export them. You can do whatever you want with them and study it in however way you want to study it.

And then, here, you have all of the different components, one by one, all the modules that had to be built to make it work. So it's quite easy to navigate this. But that's just an example of how this file was put together, so you guys can feel comfortable about it.

But going back to these folders, for instance, we took the precaution, or took the care of, if I wanted to render this animation, it should just be as easy as opening this sequence, click this button. And then we saved for all users the rendering presets so that you know exactly what we did. And there's no secret.

And there's nothing fancy going on except for these variables right here, which, like I said, if you download the project and you download the documentation, you should be able to understand why we did this. There's nothing really special about that. So we wanted to use everything as is, nothing too crazy.

But one of the nice things about this going back to the change in the process, the status quo of architectural visualization, if you will, which is something that Jaron and I, working together, struggled with for many, many years. I say struggle in a good way. We are in architectural visualization.

You have to tell a story, interpret someone's idea and story with a still image sometimes and/or with a one minute or a two-minute animation. And it just never does it the justice that the designer has in his head because this is just difficult. It's difficult to grab someone's thoughts and put them into a visual.

And then, on top of that, the designer is trying to design. And in the trying to design process, this changes daily. Daily you're changing things, and you're modifying, and you're adjusting, and you're adapting. And it was the same case with this.

Every time we did a variation of the new digital model that the designers on Safdie did, and they sent it to Neoscape. And then Neoscape sent it to our team. And as we're visualizing it, you immediately start to see these things that shouldn't be, that you thought were going to work and are not working. So you have to change and change and change. And one of the nice things that happens with this experiment is, for instance, this rendering.

In the past, we would have had to do offline rendering. And, oh, Jaron, you have to wait like a week or two for us to spit them all out. And then we'll put them on a PDF.

And then we'll have to get together in person and mark them all up. And then argue over all sorts of subjectivity, like is there a sun? The right angle of the sun is the color of the clouds. It's like the people are in the right place, whatever. In this case, we were able--

JARON LUBIN: Carlos, Carlos, not argue, debate. Debate.

CARLOS CRISTERNA: They were arguing. Debating, yes. We're debating. We're debating. But these were awesome ways to just tell the story. And together with the art direction from Ryan Cohen from Neoscape and the collaboration in between the three of us, we could just do things like that.

Say, for instance, we were not satisfied with the time of day of this image. Then we could easily do whatever we want with it and then continue to have this debate of what tells the right story for this particular space. Silly things like maybe too much fog or too little fog. Let me just undo this. Or the clouds, like that's not the right sky.

In Montreal, there aren't any clouds. All right. Well, let's turn them off. And let's see what that looks like. Or, let's add more clouds, so on and so forth. You get the point. That was one of the nice things that we wanted to chase. Was it perfect? Maybe not. There's nothing really is perfect. But it was worth trying.

And it was worth the while because of all of the things that we mentioned earlier, that we wanted to make sure that we ended up with a product that showed those ideas that Moshe had came up with back in the '60s in the best possible way we could artistically represent them with these vignettes, with this lifestyle, with this atmosphere, with the wind blowing, and the trees moving, and the water.

Like, you really are there. And I let Jaron speak for his reaction. And, also, what was Moshe's reaction when he saw it for the first time? Question to you, Jaron.

JARON LUBIN: Listen, I mean, Moshe was blown away. We were all blown away. He was touched because this is an idea that's been in his brain, not only because of the original. And it's obvious, it's like an origin story idea to see. But, also, like you said, I like your word. We're a firm that's been chasing this idea in different forms, almost a kind of a principal chase on all of our projects over 55, 60 years.

So every one of our projects aspires to achieve the effects, the environment, the kind of amenities of the original thesis of Habitat and this Hillside, in fact.

Another thing you said that was kind of triggered an idea for me, which is that all of that work is collected in our archive, which we hold. We've collected and held ideas built and unbuilt from the inception of our firm in an archive at McGill University. And so the archive becomes this total tool for us to look back at for ideas and diagrams to look forward.

In fact, this project Hillside is now being placed into our archive, which is kind of an interesting idea, which is how do you archive digital information? And we have an archivist in house, Annie Fisher, who we work with, to think about, how do you hold digital information for the long run for people to see and refer back to?

This is a topic. I mean, I'm not in the archiving industry. But that is discussed at length so that you don't lose resolution, and you don't lose data, and all that stuff. But the idea of our practice is that all ideas under the sun are here. In other words, all the things that drive our interest in creating new buildings, there's a lot of timeless principles.

So the idea of our archive is that we're always looking back to look forward. And we're excited to include this work that we've done here in the archive, looking ahead, and see what we can do. I mean. I'm personally interested in seeing-- because this model is out there for downloading and for people to play with-- what versions of Habitat will other people conceive of?

I imagine a super savvy young kid making a model just like Moshe made when he was in his 20s. Or maybe now, with how familiar people are with computer modeling, like what if some 12-year-old has access to this? Because it's like, some of the tools here are not that easy. But they're pretty accessible.

So people learning about and playing around with these tools, I think, generally can bring architecture, and architectural thinking, and spatial thinking to a larger audience much earlier. So you don't need to be going to a high-level graduate school to be learning these things but that we can get these out there more.

The archive, by the way, our archive is open and accessible to anyone with access online. So you can not only download this from Epic but you can have access to all the work that we've done over the legacy of the practice online.

CARLOS CRISTERNA: When we started, one of the more interesting things for me was like that model we see on the left, that photo is obviously an old photograph. And then we see this digital version of the model that we received, Neoscape received this model. And then we embellished it a little bit more in Unreal Engine.

But I've seen this model in person. I've seen this model at Safdie Architects' offices, as well as these sketches. And we figured, for this presentation, let's try to find that parallel of these images. And see if we can find it because they're obviously not an exact design, one to one. These were sketches ideas that Moshe drew back in the day.

But we tried the best we could to see what would those sketches look like digitally and compare side by side like this one right here, over there. Or this one over here, something like that.

But then the ones that Jaron was referring to, which was very much interesting, was this one. This is the actual, physical model of the old-time photos. The old Black and white photos and all the photos from the '60s, this is the model. And this was maybe a year ago, we were there in Boston. And I just briefly snapped these photos with my phone.

And then we just did a couple of screenshots of, generally, the same view just to see what that would look like, more to scale than the massing model itself. Once you have all the vegetation and the tiny little vehicles and you know furniture and things like that, it really puts things into context a bit better than, obviously, a conceptual-massing model.

It's a massive orthographic view of the project.

JARON LUBIN: Yeah, I mean these really-- the drone photos that you shared with us were kind of mind blowing, also, to know that these were kind of generated from a real-time model that we were flying around together. So like the idea that it's not a still anymore but it's an environment is like a paradigm change for us.

And we see a lot of students at architecture schools in the area using tools like this already in the curriculums and moving away from that previous model of, all right. Frame a view. Let's perfectly create it, and maybe moving more towards these production of environments in total.

So there's a huge change that's-- I mean, it's been in the works. But we're seeing it really come through. And this technology really is like upper level, showing the potential of it.

CARLOS CRISTERNA: Yeah, I mean, it's amazing the-- technically speaking and putting a little bit aside-- the value of the design itself, just the fact that we are literally going through the clouds in a real-time model.

And you can see all of that, of the city, with the project right there. I mean, to me, it is mind blowing. And I hope it's fun. I hope people get a chance to download it, and learn from, it and shoot at us any question they may have and enjoy the work that we've done for the last couple of years. And it's useful to people.

JARON LUBIN: Carlos, thanks for joining. I mean, we're both looking forward to doing this presentation again in person. But hopefully this video makes it out to everybody who can and cannot make it to our little panel discussion.

CARLOS CRISTERNA: Well, thank you very much for inviting me over, Jaron, and for partnering with us. And a shout out to Neoscape, as well, for being a good partner with us and helping us produce such a nice digital archive of a great idea that we hope many other designers get inspired by.