CDFAM Amsterdam 2025 · Amsterdam · 9–10 July 2025
Democratising Computational Design via Cloud Based Applications
Abstract
While computational designers have always created powerful and sophisticated software, the reach of their work often couldn’t exceed that of their modeling tool of choice. We’ll show a novel way to build and deploy web applications, including complex user interfaces and interactions with 3d geometry and data visualization with nothing but Grasshopper.
As part of ShapeDiver’s presentation, Age Van Boxem will present Orthosolid, a platform powered by ShapeDiver that enables clinicians to design and order custom 3D-printed orthoses through an intuitive digital interface. He will explore how computational design and UX come together to simplify complex and outdated workflows, making advanced customization accessible within clinical practice and without engineering expertise.
Transcript
From YouTube’s automatic captions, lightly cleaned; expect some errors. Each timestamp opens the video at that moment.
Read the full transcript · 2,880 words
Okay. Hello everyone. I am Edwin Hernandez. I am the head of computational design and projects at ShapeDiver. And today I want to talk to you about how to democratize computational design via cloud-based applications. As many of you know, one of the main softwares to create computational design nowadays or the most known ones is Grasshopper. So I want to talk to you about how to democratize computational design through Grasshopper and through well cloud-based applications.
0:31 Grasshopper has been around since 2007. But I got to know Grasshopper until 2014 when I was doing my bachelor’s in architecture in Bgotaa Colombia, my hometown. Yesterday there was many Latin Americans here as well. So that was very nice to see so many Latinos in the community. And back then my teacher was from Canada, so the class was in English and I didn’t have a very good English.
1:05 My English was not very good-looking. So but even though my English was not very good-looking I was able to manage to see the potential of Grasshopper and managed to do some of these little projects in 2014 when I was introduced to it where we did even some a little bit of digital fabrication where we were like unwrapping and we were nesting and then we were doing the laser cutting.
1:30 So even though I had this English problem, I was really able to see the potential of Grasshopper. But the year later, I decided I wanted to fix that problem. So I went to Australia and then I spent there two years. I was not just learning English, but I was also doing a bachelor in computer animation. So here is just a little thing that I did for animation purposes.
2:00 And I was also doing some freelancing for housing especially doing some drafting some 3D modeling and finally I was also teaching some architectural students archicat to pay my bills I was mainly doing the housing part doing those drafting and 3D modeling however that’s has a lot of computation competition there so I was not able to do much money there so I had to specialize in something else and there is where I start to get a lot of clients in doing the sheds like this big like the place that we are right now places where normally is used for farming or or like industry and well in Australia there is a lot of space so they have a lot of farming so I start to get clients more in this kind of industry not just in Australia but also in New Zealand and with this I was able to automate a bit the process my my process of creating these drawings because this kind of these kind of buildings are very standardized so I had some templates.
3:00 I was able to optimize my processes and to make more profit from it. But then I started to see that well if they are so standardized then I could potentially do even more automation and to give somehow a way of the clients themselves to easily change their shed shade their parame change the parameters of the of the shed. So I started to look into solutions to do that and well the the best way to do that is via a browser.
3:24 So when I was doing computational sorry animation so computational animation then I learned this this tool like cinema 4D blender 3dxmax all of these are mainly for animation and gaming and then I also was introduced to something called sketch where you can host your 3D models so that anybody can interact with them. So it’s like before we had MP4 files, we had our videos in our computers locally but we were not able to share them to everyone but now when YouTube appeared now we were able to do that.
4:00 So Sketch Fab is the same than YouTube but for 3D models and through their API I was able to create different kind of configurators where I was able to change materials. I was able to do even a bit of modular, hiding, showing things and even faking a little bit of parame of being parametric by like overlapping objects or stretching objects, things like that. However, well, this is done with explicit modeling.
4:30 So, basically, I have to actually model the objects and it’s just one at a time. So, it’s not very it’s not really parametric. No, I’m just faking it. So, I started to see the bottleneck and I wanted to make this even more powerful. So in 2017 I was like there’s got to be a better way to do this. I want to make tools that are more powerful and I also want to even extract more information from these 3D models not just visualize them but actually create some processes from them some maybe structural analysis etc.
4:58 So I already knew a sketch fab and I remembered that while I was studying architecture I was introduced to Grasshopper but then I was missing that link to the web. No, we have a sketch of our 3D models for YouTube for videos and then what do we have for Grasshopper? So after a little bit of research then I found out that well that already exists and that was shape diaper.
5:24 And well that was in 2017. So how does how do we convert this the spaghetti monster of Grasshopper into the actual 3D application the an application that anybody can access through a browser and u well they don’t have to deal with the complexity of the grasshopper model itself. So here we have our computational designer in their local machine and now we need to put that into a server to begin with.
6:00 No, so that anybody can access this model and how do we do that? So in 2019 magn introduced Rhino compute and that basically helps you do that run grasshopper in the servers. However, shape diaper has been around since 2016 and you could say in 2015 when the idea was born. So well, basically we we didn’t have Rhino compute. So through all of these years, now 10 years, we’ve been able to figure out the best way to run Grasshopper in a server.
6:34 And for example things like what happens if you have hundreds and hundreds of users that want to use the Grasshopper script the the the actual final application and how do you do to serve each of them efficiently if they want to use for example the same the same model or what happens if they want to use multiple instances of Grasshopper. How do you dispatch that so that you can scale up your servers?
7:02 You can scale up the use of Grasshopper without without overwhelming overblowing your your your service. No, you if what if do you have to initialize one instance of Grasshopper every single time that somebody wants to access it? Well, no. You that that will overblow your server. So that’s why we have an algorithm where we have the Q in theory and then we are able to really dispatch efficiently and at scale the use of Grasshopper.
7:29 Well now we are able to run grassina server but now we also need to display the geometry in the browser. No in in Rhino you normally have your rhino viewport but now we need to display the geometry but in a website. So how do we do that? Now we have something called webjl. This technology allows you to do that to to see the 3D models in in a browser.
7:53 And on top of that, you can use these frameworks like BabylonJS or 3GS which allows you to map then the geometry that you normally see in in Rhino map it into geometry that you can see in a browser. In our case we use 3GS. So now you can run your grass personal service you can display the geometry results but now we need to give access to the parameters create a user interface.
8:23 So all of the sliders, dropown menus that you normally see in Grasshopper, how do you expose that in the browser? So in one route you would use web frameworks, so regular CSS, HTML, JavaScript to create all of your sliders, all of your drop-down menus, map what you get from Grasshopper into the browser. Or on the other hand, if you want to stay within the within the computational design field and you don’t want to start coding then you can use our plug-in the ShapeDiver plug-in which has a section called app builder where you can not just so with grass you cannot just design what the user will see the final result the final calculations but also the user interface itself.
9:02 So you can decide okay I want this slider to appear in the left side menu, the right side menu, the left side or I want this exact parameter to to to be shown when X or Y happens. So conditional and all of that through Grasshopper and also whether you want to show tables with information, whether you want to show graphs etc. All of that can be done only by using your Grasshopper skills.
9:31 Okay now we can we can run grass operina server we can display the geometry results we have the we have given access to the parameters via user interface but now that we are already in the cloud we’re already in the browser so to make this even more powerful we can start to connect to other services for example via APIs and we have our own API so you can do do you can do that connection to these external services so we have a a viewer API, a backend API, so you can connect directly to our servers, a platform API, and then you could connect potentially to any of these services.
But if you don’t want to also do some coding, but you want to stay again within your computational design skills, then you can use also some plugins that we have already developed. So so that you don’t have to do any coding at all. So for example, we have an integration with Shopify or WooCommerce if you are into consumer products. So if you sell rings, furniture, these kind of things and you’re a great computational designer and then you want to sell those products, you can do it directly just by using our Shopify or WooCommerce integration.
10:40 Or what if you are in the AC industry? Then well many of you may know speckle where you can then from shape diver push information into speckle and then from a spec push it to revit or push it to techloud or any other of these AC software or even an integration with open AI so you can start using some artificial intelligence with your models. How do we use it in this artificial intelligence?
11:06 So here I have a video where okay you’re creating the UI and everything in Grasshopper. Then you have this ring that you then connect with just a an ID to our Shopify plug-in and then here we’re already in Shopify. And then all of these parameters were were all defined in Grasshopper. So the user can either start to click on the parameters in Grasshopper sorry in the in the user interface or there is a second option where you can have like a chat bot that is basically open AAI in the background and you can start to request u changes in your model.
11:40 See it’s going to happen in a bit and yeah there we go. And then you can write okay I prefer X metal or whatever. And then you have the context of of all of your parameters because they come from Grasshopper. Plus you can give extra context through through through Grasshopper. So Shape Diver is u works with 12 plus different industries worldwide and some of those industries special or not they use ID manufacturing.
12:13 So for example, eyewear, footwear, for example, for footwear, we have Bivo barefoot and bivo barefoot for example, use shape diver to help help accelerate the process of generating the best last. And here we have a small video, one of the latest products that they are releasing where they are actually creating the entire the entire shoe 3D printed and they use also ShapeDiver to help the user do some final design decisions.
12:43 In the process. Then we have Saib in concrete 3D printing. So here for example is an example in the Netherlands which they use concrete 3D printing and in this case they also use ShapeDiver well to give access to the grasshopper files so that people can do some design decisions in on their own as well. And then we have auto solid which is here with us today. And then they will talk us they will tell us about how they use shape diver.
13:20 Thank you Edwin. So let me start off by introducing you to Mr. Val over here. He’s an orthopedic technician in Belgium. And he creates custom orthopedic devices for his patients. How he works is he sees many patients let’s say in the morning and then afterwards he creates all his products inside this atalier. If he wishes to expand his business or grow his business he would have to expand his atalier capabilities which is very expensive and also find new personnel to create these devices manually.
14:02 Even if he does those two things, he finds that new personnel. Traditional methods of creating orthopedic devices are still very manual and quite repetitive and thus difficult to scale up. And he’s not alone in Belgium. The whole sector in Belgium and I’ve also gathered information from the Netherlands is that demand is increasing for orthopedic devices but personnel availability qualified personnel availability is actually slowing down all that while the work that they do remains really repetitive and thus difficult to scale and so is able to fill this void by offering by completely revamping their traditional workflow.
14:51 We offer modern measurement techniques like 3D scanning. We of course offer modern and future proof production product creation tools. And then we combine those with modern production techniques, i.e. 3D printing in our case. And so all this is sort of gathered and compiled by our computational models or parametric models which we have in Grasshopper and bring to life with shape. Now this might sound all a little vague.
15:24 So just let me show you how we make a little oh to play it. Let me show you how a 3D printed thumb speaker is made in our platform. So as you can see some data points are placed and trim lines can be adjusted really easily by dragging the points and defining the trim lines. After this the brace is generated generated automatically and our orthopedic technician is able to choose a bunch of options to fine-tune the design for their patient.
16:05 They see a final preview of the product and then they order inside of our web application. And so as you saw in the video, clients are now able to semi-automatically create these devices and they’re the same or even better quality and they are able to do this in a shorter amount of time. Also investments to start are kept really low both in terms of the learning curve which is really important in this sector and also financially also really important as I’ve talked about before and so this offers clients like Ferala a really interesting opportunity because now they can scale their business without big investments nor compromising on the quality of these devices and So this workflow also offers us as a business really important advantages.
17:08 Scalability being the most business advantage. But for me as a designer I have both control over the parametric model inside of Grasshopper and maybe even more importantly I have control over the actual UI and UIX that we have in the platform all in one human. And this ease of developing these tools is really important because this means that we can also evolve or solid over time and so we can grow together with our clients as they grow in expertise as well since this is really an emerging market.
17:46 And so we try to try to strike that balance between the control that we offer in our Grasshopper models and the ease of which we can deliver this to our clients. And so I hope you can see where I’m getting to. Offering these computational or parametric models or designs directly to end consumers or in this case orthopedic technicians is is not so much about how complex or advanced I can make my model.
18:24 It’s about how we can make these tools accessible to people who use them to help their patients. And that’s their ultimate goal and also our ultimate goal. Thanks. To learn more about the CDFAM computational design symposium series, to see the archives of previous presentations, and to learn about future events, visit CDFAM.com.
More from CDFAM Amsterdam 2025

Flexible Geometric Modeling and Atypical Simulation Solvers to Streamline Design Optimization
Wesley Essink · Altair

Design Optimization for Advanced Manufacturing through Forward Looking Performance Simulation
Chris Robinson · ANSYS

AI Judges In Design: Statistical Perspectives On Achieving Human Expert Equivalence With VLMs
Kristen Edwards · MIT

From 2D to Mass Production: Computational Design at Scale with Toolkit3D
Sarah Clevinger · Toolkit3D

From Days to Hours – Accelerating the RFQ process through scalable FEA automations
Andrew Sartorelli · Synera

Real-Time Computer-Aided Optimization (CAO): How GPU-Native CFD Changes the Industry
Qiqi Wang; Momchil Minkov · FlexCompute

Geodesic Slicing: A Generalised Framework for Multi-Axis 3D Printing
Alessandro Zomparelli · AIBuild

Stress-based Design of Lightweight Horizontal Structures for 3D Concrete Printing
Luca Breseghello · DTU

Building Surrogate models for Physics Simulation using a no-code approach
Asparuh Stoyanov · Key-Ward
















