0:00 My name is Rick, I’m a senior computational designer and structural engineer at Arup in Amsterdam, and today I’m going to tell you a little story about a few projects where we’ve used computational design in the built environment. Times are changing, we have fa, we are facing very important challenges globally, we have climate change, housing shortage, and people are moving into the cities, which is leading to densification, very complex sites, as well as hot, dark and windy environments which we have to design for. At the same time, we want to design for a sustainable world, we need to think about how we produce and use our energy, reduce our material use, and improve the biodiversity. So how can we actually design comfortably and healthy and a sustainable environment where people can enjoy the sun in a nice environment with in the build environment, while also controlling the design process and managing risk.

1:00 Therefore, we need to go back a few centuries to Vitrus, because he wrote in one of his books about the three principles of design, and historically it can be observed that the dominant stakeholder in the design process will dominate the outcome. This can be seen in the 12th century, U, in the gothic churches, where the beauty, the the venas, was the most important stakeholder, then in the industrial revolution the firmness, the strength, was most important, so they want to show off the construction material, the production techniques, and later on the util us. We’ve used, we’ve heard it a few times throughout this conference, form follows function is still very important architectural term, but how and if we really want to design for health and comfort within the build environment, we should not only think about form foros function, but we should also design for our people and planet, and therefore we believe in a paradigm shift which says form follows people and planet.

2:13 So one wise man said once, if we do all of our exploring within our own heads, we tend to be limited by what we already know, and with the use of computational design we can actually start to explore more designs and include these KPIs for health, comfort and sustainability in the earli phases of the design, because there we think there’s the most power to include these techniques, where you have the most ability to still impact the design while the cost is still relatively low. And at Arab we believe in a very holistic design approach, where we follow this integral process, where we discuss client ambitions UPF front, that’s the most important one, we need to define our boundary conditions and our variables, with that we we can determine our logic which we use to generate all these alternatives, which can be buildings, master plans, or building components.

3:07 Over the years we’ve built up a few analysis modules which we then link to these options that we’ve generated, and then we can start to explore and evaluate and optimize this, this is all running through the cloud, we’ve heard this before, but I’ll show you some examples later on, and in the end we want to present it in a way that the client and and architect can interact with it and actually make datadriven decisions. So my story started in 2016, when I joined Arab as a structural engineer, and I didn’t know what parametric design or computational design was, so my colleagues were working on this project in Amsterdam where they were doing automated daylight calculations with the architect MVRV, so every time they changed the geometry we had a a script running to to calculate the day, and I thought, okay, this is something really interesting and I want to learn more about it.

4:03 So I got the opportunity to work on a project in the UT Utra City, very close to the central station, where the competition was about generating 20,000 square meter GFA in a very complex site, so we had to convert an an area into a peaceful urban oasis, where the residents of Utre really wanted to have a sunny park, a good comfort, less sound, this resulted in a competition winning design. And that led me to the coffee machine in 2019, where my manager approached me, Rick, do you want to work abroad, and I said, yeah, it depends, and then he said, do you want to work on a stadium, and I I said, okay, yes, I still had to convince my girlfriend, but I already had said yes.

4:56 So I went went on the plane to Madrid, our our Madrid office, to work on the Barcelona Stadium, so that was really really nice, so I worked there for a year on this stadium, but as the front page of my presentation showed already, that’s not what I’m going to talk about, because after a few weeks, also there at the coffee machine, I met a guy from the facade team, and they said can you help us out with the facade of another stadium, and that’s what I’m going to talk about. So the the this is the Real Madrid stadium in in Madrid, of course, where Arab provided the facade engineering and the lighting engineering of this very unique iconic stadium. So the first construction started in 1944, and currently there are, they have a capacity of 84,000, you can see the very iconic facade, but another really nice feature that they have is the retractable roof and the retractable pitch, so if you have time after this conference, or tonight you watch the Champions League game, maybe you can see some of that.

6:05 So the design was made by GMP and L35, a very complex and unique shape with a reflection of the facade, and we had to model this, we had to analyze this, and and and maybe also come with options to explore the construction. So when I went to Madrid in 2019 it looked like this is, and I went there a few months ago and it’s nearly finished, so I had a nice task to make a computational design model of the whole facade, more than 13,000 panels, each having different configurations, different angles, and like we all love color images, me too, so we did some analysis on the different angle configurations of the different panels to see if we could impact construction techniques or make it more easy to fabricate.

6:56 But the client was most interested in something else, and I didn’t know the previous presentation was going to touch on that as well, but he was worried about this story from the tower in London about the reflections frying a car, you’ve heard the story just now from the glare, so we also had to do a glare analysis, because yeah, the the client was really afraid that we would fry the buildings, or that we would have reflections on the roads next to the stadium. I’m sorry Gabriel, we did do the the bouncing, but that’s the next slide. So what we had to come up with is a system of changing the facade in order to reduce the reflections, so we were able to influence the architect by making slight changes to the top louver in order to push the reflections back up, but we also had to consider if we change it too much maybe the reflections come back from the other panel.

So we basically analyze a lot of different panels, a lot of different sunrays, in order to get some insight into this, a lot of computation power, in the end we were able to sort of translate a change of facade rotation into the number of hits to the to the environment, so the number of hits to the roads or to the bu, neighboring buildings, but then you can imagine if we change it a lot the architect will fire us because they don’t like that we change their design too much. So we have seen this Paro front as well, this is definitely a Paro front example, because we can change the facade a lot and reduce the number of hits to the environment of the number of reflections, or we accept a little bit more h hits and change the reflection properties of the material, and therefore we we did some studies.

So what if we don’t accept a lot of hits, the red means that we need to change the facade a lot, and the architect probably would not like that, so we could also shift a bit, so we accept a little bit more reflection or a little bit more, and you can see slightly the facade turning back to the original shape. Then from all of these results we mapped some Gan functions to still have a nice and fluent iconic shape, so that we could actually present something U that is that could be produced to the client and to the architect. But how could we actually inform the design team, this was our very important question, because you can do all of these analysis, but you still have to convince the client and the architect to go on with your design and and and accept the decisions that you want to make.

9:45 So thanks to McNeil, we were using Rhino Compute, or we were testing Rhino Compute back then, to actually send a model to the client and let them play with the sliders, let them play with the change of the shape, and see how it would actually affect the reflections of the facade, and it was really crucial in our development of a tool which is called Arab Inform, and I’ll tell you something more about it in the next projects. Unfortunately we didn’t have enough time to really launch this and and share this with the client yet, and in Madrid we had a Unity team team that could actually push our results very quickly into a Unity viewer, the game engine, and they could actually explore the different configurations and and see how the the change of shape would actually respond to the reflections at the at the different locations around the stadium.

10:42 This is me a few months ago at a rainy day in Madrid, it was really nice to to visit the stadium finally when it’s almost finished, I actually got this to go on the pitch as well, which was really really cool, so you can imagine what to do after this project. So then while I was there Corona came, and I decided to go back to my girlfriend because she was a bit sad that I was gone for such a long time, so I went back to the Netherlands. I got the opportunity to work on a very nice project in Amsterdam which is called Elements, again a competition, we were collaborating with Kuk Architects from the Netherlands, and from the from the start of the project we said, okay, let’s do something different, let’s create a shape of the tower that follows the people and the planet.

11:36 So we agreed some KPIs on the building scale, on the envir, on the neighborhood scale, and on the planet scale, therefore we assessed daylight, sun, wind, we included the green onto the building, strategically position PV panels, and also try to reduce the embodied carbon. So what we then did is we did a lot of exploration together with the client and the architect, we came up with some variables, how can we play with the shape of the building, and this led to a lot of configurations, but also making the informed decision on all of the data that we generated, and that led to the competition winning design which you see here, a render of which we also included some additional features.

12:21 We wanted to reduce the the embodied carbon, so we introduced quite some Timber structure in there, we analyzed the wind conditions next to the balconies, that actually helped a lot in reducing the wind impact on the ground level, and on the roof levels, and on the balconies we could position the PV panels to actually have a good PV coverage at the peak hours. In the end we achieved more than a 50% reduction on embodied carbon and operational carbon, and here are some images of the top of the tower, and I I got asked a question just now, are we going to see some construction side images, this is from a few weeks ago, so we’re actually constructing this tower as we speak, hopefully next year we’ll see this nice shaped by the elements tower in Amsterdam.

13:12 So after this, while I’m still working on this project, I got to help out on my final project, which I’m going to share with you today, is a high-rise tower in rdam where the client wanted a 60,000 square met tower, they didn’t know the shape, they didn’t know the height, they didn’t know the location, they just knew the plot, and that was very interesting as a computational designer, because then you can actually start to navigate a complex site, explore configurations, and they helped us with defining the ambitions. So they said, okay, we want 60,000 square met GVA, we can explore the sun on the building for PV, but also on squares in the neighborhood in rdam, because the Netherlands, the Dutch people love the sun, we were building very close to neighboring buildings, so we had to assess daylight, we had to assess noise, structures.

14:01 We were building very close to a Metro tunnel, so we built a component that could actually predict the differential settlements when we would construct the high-rise tower next to the the Metro tunnel, and finally we did also do some wind comfort studies, and on this topic we have been doing research in the past three years where we wanted to up the speed of our calculations, because a CFD model for one of these projects would take us about four hours per model. So we’ve developed a machine learning algorithm to quickly assess the wind comfort when we position an urban environment, so we could actually predict the wind comfort in a in for 11,000 options in one and a half hour, versus 5 years, and here you can see it in action at a fair where we also built some Lego buildings, and you could play around, people could see the the the effect of positioning different buildings and heights on the wind comfort.

15:12 So with that we explored a lot of options on different KPIs, taking into account the settlements, taking into account the sun, and all the the the KPIs that I showed earlier, and that we pushed into our parameter space, you might know the Design Explorer from Torum Tomasetti, we we also created our own, so we could actually explore the data where each line represents an option, so what’s nice about it, you can actually play with the output to get the input. So therefore, what if I want to minimize my settlements, I want to maximize my sun, I want to maximize my daylight, what could be an interesting option, or the client said, okay, I want to maximize my GFA, I want to maximize my daylight for the neighboring building, what could be a nice option.

16:01 So we actually shared this with the client and with the architect, and we came to three options which we then presented into our Arab Inform tool, which is hosted in the cloud, running Rhino Compute, which we can then use to inform our team on how, if we change the design, how it would affect the different KPIs, such as wind, such as settlements, and that led to the final design, which werere now in to the preliminary design. What was really nice on all these projects is that the client and the architects really enjoyed this data driven design process, and I found a quote once by Parametric Monkey, solutions that we generate should should be seen as a starting point to underlay, trace over, modify, to actually create a better solution.

16:52 Because with computational design we can explore more options, we can design efficiently and quantifiably better, and most importantly also collaboratively find the best fit compromise solution, in order to design for form follows our planet. Thank you.