Innovative Generative Design in Revit and Civil 3d for a sustainable city

PRESENTER: So hello, everybody. Thank you for coming for my presentation at AU 22. The topic is about "Innovative Generative Design in Revit and Civil 3D for a Sustainable City."

My name is [INAUDIBLE] Marechal. I'm head of city information modeling department at COLAS Group. And my core speaker, which unfortunately is not live, is Romeo Sanchez from Autodesk. So first, a few words about my company. My company is called COLAS. It's an international group.

And it's the leader of construction of infrastructure projects and maintenance of infrastructure projects and also looking forward to create sustainable mobility. A few numbers on my company-- we have 57,000 employees working on the five continents in 50 countries, doing 13 billion euros in revenue and working with 8 million tons of recycled materials on the last year.

We have a dense international networks of asphalt plants, quarries, concrete plants, recycling units, but also subsidiaries. And maybe you don't know us as COLAS in the US, but we have different brands, like Delta, Reeves, Simon, Barrett, or Sully-Miller in the USA. And in Canada, we are called MAC Asphalt or Sintra, Terus. So we have different brands in all over the world. So we are an international company.

And we work on a wide variety of projects, mainly on roads, highways, civil engineering, airports runways, ports, urban developments, tramways, bus rapid transit systems, bike paths, and shopping centers, mainly. We also work on buildings, actually. And we have a strong focus on innovation, on ecological design. So we have different products.

And one of them is called Flowell. And I will focus just right after that on that specific innovation product. So Flowell, let me show you a small video about it so it will be clearer to understand what we're talking about.

[VIDEO PLAYBACK]

- In the evening, especially when it gets dark early, I don't feel safe when I walk home. I can tell the drivers have trouble seeing me. The crosswalk is supposed to be the safest way to cross the street, but the white stripes are not easy to see in the dark.

We should invent a luminous crosswalk. We could see our children in the distance crossing the street, with the [AUDIO OUT] lighting up under their feet. COALS' engineers are hard at work to make roads safer. They have invented a solution to make crosswalks visible at night.

The light panels are very thin and very resistant. Vehicles can drive on it without skidding. Imagine all the possibilities offered by these light strips. They can create dynamic [AUDIO OUT] on command. They can take different shapes, colors. [INAUDIBLE] dotted lines indicate directions, warn of danger, secure bus stops, and more. Tonight, a new road is ahead of us-- safer for both pedestrians and drivers. There are so many ways to make the world brighter.

[END PLAYBACK]

PRESENTER: So that is the product. Here are a few examples of projects that we've been working on. So at the top left, it's project in [INAUDIBLE] France for bus crossroads. At the bottom left, it's for crossroad. And at the right, it's for bike pass.

But these are real projects. So as you can understand, it's LEDs encrusted into the roads. And then we can make different way to light and to display.

So we've been working with design to make sustainable cities. When we talk about sustainable cities, because the aim of Flowell is-- as you can understand, the flow is to-- to make the flow better into cities. For example, we could change a road with three ways from two plus one in one direction in the morning. And then in the evening we can change it thanks to the light applied on the roads.

And also it's been used at first to make the city safer because our first market is crossroads for pedestrians. Actually, we could do almost everything with this, any form, any colors, any lights. But we try to focus on system which can be repeatable into the city because, of course, it would be cheaper. So the way that we think about its development is really to find a use case that will be repeatable.

So the project on which I want to focus today is called La Defense in Paris. It's a big project for Flowell because we had more than 100 LEDs encrusted into the roadworks. And the problem was to find the best way possible to connect them. So that's why we used a generative design on this project.

So maybe a video to present to the project. So it's in Paris, La Defense. As you can see, you have some lines on the roads, some straight lines. And then it's a switch off, and you have some curved lines.

It's synchronized with, how we call it, with traffic lights. It's synchronized with traffic lights. So if it's green, it's in a one way. If it's red, it's in another way. And for pedestrian, it's synchronized with detection. So if you are detected, if the presence of a pedestrian is detected, the light goes on.

So as you can see, we have different system where the crossroads-- we have three crossroads on these projects. So this is-- there's a real situation, how it looks like. So we have three crossroads, and also we have the line in the middle of the road for the car.

So the curb lines, as you can see right now, and then the straight lines in this way. So as I said, it's a lot of ridges to be connected. And we have a lot of possibilities to find the best way to connect it.

So we've been using generative design. Actually, it's almost the same on electric components. It looks like the same. The road becomes an electric component on which we are going to create pass to connect the different electric system between them.

So we had the look with Autodesk Consulting at first to see what was the best product to do that generative design. It could have been pure electronic products like-- I don't actually-- I don't remember the name. But it could have been a really specific product, not Revit or something else.

But at last, we decided to do it on Revit and generative design. Well, at that time, it was called Refinery. So the step for this project was to define what were the problems to solve and to integrate the constraint of the sites.

So the problem is really to find the best configuration to connect all the LEDs, but also the junction boxes. So it's about how we are going to group objects and how we are going to place it on the project. Then is what size, which it implies for the wires. So it's length. It's the diameters, but also the trenches into the road, so the widths and the depths.

In regard of the constraints of the sites, because we cannot do whatever we want, so we have some constraints or some areas on which we cannot go because we have some existing networks and also, of course, the costs, which will be in place. And at last, the goal is to simplify the access and the maintenance of that kind of systems.

So how do we seem to specify a specific program in generative design? We have, as you can see on the right, it's two different solution on really simple example of what could have been done on that project. So we have picked up only six LEDs.

On our project, we have, as I said, 100. And the idea is to create the different variables on which we want to work on. So the variables are in red. So it can be each one extra. And the idea is really to define the inputs, so that variables and which range we want to address.

And then for each inputs, we have different attributes like what kind of object it is, what's their position, and other information. I go at the end of the slides-- the output, of course, is what we are going to look at to define our best solutions.

So in our case, it was the trenches. So their lengths, their heights, so the number of objects. And between the input and the output, of course, we are z algorithm. So that was Dynamo script, Python scripts that we ran through CV 3D and Revit.

So that part is really tricky. It can take times. And it's really the heart of the problem, really to define it because you need to understand your project. So we had electric engineer. We had someone we know who knew the project.

Needs global view. We had a developer-specific coder from Autodesk, actually, Paolo Serra. So you need, really, a lot of different skills. And it can be really tricky and take times. And that part should be actually detailed by my co-speaker, which is not here today.

But it was about how design optimization works. So I go fast on that one. And I'll let you get contact with me if you want to know more about that. So how will we process the solution of this project?

First of all, we use Civil 3D to create the project, to place LEDs, to place the electrical cabinets, to place exclusion zones, and junction boxes zones. So actually, it looks like this. So it's Civil 3D, or it can be AutoCAD.

But I show you through a Civil 3D because we needed to use Dynamo on it. So Dynamo was used to detect all that objects, their categories, their properties, and to transform them into a data file. Actually, it was JSON file.

And then it looked like this. So you can see different areas. It can be exclusion zone, which can be LED objects. And then we compute it into generative design with Revit and with specific Python scripts to create hundreds of solutions to connect the LEDs, to locate the different junction boxes, and to dimension the trenches.

And this is how it looked like in a generative design or Refinery at the time. And this is one of the solutions that we picked up. So as you can see, we have to make connection between all the LEDs and then to connect the different group of LEDs into injection boxes, which has the blue points. And all the blue point have to be linked with the electrical cabinet, which is at the top of these images.

And at last, we use Revit to create and place automatically the Revit families for wires into the trenches because once you've been picking up the best solution, it's only lines, pulley lines, like you can see in red at the top left. But once you have chosen the right path you want to use, it trace automatically the Revit families, so with the diameters and with the 3D geometry.

So it looks like this, actually. So you can filter the different solutions. So you can have hundreds of solution. You picked up one. You filter it, and then you applied the script, which creates automatically the different objects into rates. Like that.

So this is different trenches. The filter is used to define the different output that you want to work on and to get the best solutions. Then we could go a bit deeper into the code, the Python code. Actually, that presentation is not finished yet. So it will be live on AU 22 if you are lucky enough to be there.

If not, just contact me to know more about how we did the code on Python and Dynamo. And I will explain to you. So thank you for listening to me.

Feel free to contact me if you want to know more about the project, about the product, about how we use generative design and Autodesk project. And thank you, and see you back.