설명
주요 학습
- Explore several business objectives that can be affected with generative design.
- Watch several iterations of the Generative Design study setup and solutions.
- Learn how to determine which outcomes meet various objectives using the Explore tool.
- Witness the real-world outcomes achieved by Britten in this project and others.
발표자
- RJRyan JohnsonI lead the Team D3 Education and Support teams as part of our Client Success organization. I'm responsible for the design, implementation, and delivery of Team D3 client education & support programs. To accomplish this, I build and maintain the tools and programs which we use to deliver measurable, validated results to our clients, and I guide and direct the teams who deliver services through these systems. I'm also proud to manage the development of CAD LIVE, Team D3's unique Live Instructor-Led Virtual Education platform which we use to deliver live virtual training across the globe. We inspire clients with the skills and tools to build a better world, led by our continuously educated teams!
RYAN JOHNSON: Well, welcome everybody to Generative Design-- More Than Just a Buzzword. This is a case study. My name is Ryan Johnson, and we have with us also Isaac Thelander. We'll do a few introductions here in just a minute. But welcome to Generative Design-- More Than Just a Buzzword.
So you're probably expecting a few objectives based on what we said we were going to teach you in this class. We've got four main points here. Number 1, exploring several business objectives that can be affected with generative design inside of Fusion 360. Number 2, watch how some iterations of the study setup and solutions work. Number 3, using the explore tool to comb through all the outcomes you get. And number 4, witnessing some real-world outcomes from these tools. So that's what we're going to show you today.
First, a few quick stats about our projects. So we did a project with Britten, who's joining me for this class. 559 outcomes generated allegedly with generative design. 11 studies total. That was between Team D3 and Britten.
Eight prototypes destroyed by Isaac, who loves to destroy prototypes. You'll hear from him later on about that. And don't know, Isaac, is this accurate-- 120 hours saved?
ISAAC THELANDER: I think 120 hours or more. Honestly, it's one of my favorite things about generative design. I got to put in just the very basics of what I need and let the computer do the rest. And maybe a little bit more than that, but you'll see.
RYAN JOHNSON: Awesome. Awesome. We will see. Thanks for that. OK.
So before we get started, let's go ahead and meet the teams. And so, I'm going to first turn it back over to Isaac to introduce himself.
ISAAC THELANDER: Hi, I'm Isaac Thelander. I'm a Product Engineer here at Britten. Pretty shortly out of college-- so what that means is I get to experience all of the newest and coolest softwares that the other engineers don't really want to learn right now. But that also means that I get to be the first person to have hands-on experience with Fusion 360's new features. Like, in this case, generative design, a shiny new toy that Ryan showed me and I was able to run with and play with.
As for Britten, this is a little bit of a long intro. You can read about it if you want. Britten, we are a marketing company. Anything you can think of that you want to do-- big signage, big displays, small signage, small displays-- anything you want to do to market your company, we can come up with a way to do it. We make shipping container displays for trade shows, for bars, for interactive experiences, giant foam sculptures custom painted, like the big cave that you see there in the bottom, building wraps around anything that has concrete that you can attach to, or just custom fixturing for a retail shop, or pyrotechnics for the Denver Broncos. We're very proud of that one.
RYAN JOHNSON: Awesome. Thanks for the intro, Isaac. OK. So moving on to myself, I'm Ryan Johnson. I'm Manager of Education Support here at Team D3. I've used Fusion 360 since the beginning.
In my 15 years with Team D3, I've mostly spent on the technical side doing actual consulting work, training, tech support, those types of things. But for the last three years, I've been able to help manage those services for our clients like Britten as part of our client success team specifically. So a little bit about Team D3, we're an Autodesk Platinum Partner in North America handling various industries across the country.
And we really pride ourselves on our services over anything else-- our technical services that we offer. And I mentioned client success a little bit ago. I'm part of our client success team, which I think is great. Because, as you see at the top, our mission statement as a company-- we only succeed when our clients succeed, and get to help them do that, as hopefully this case study will show you.
So we're going to go back to our objectives here-- the four main objectives that you're hoping to get out of this class. But we're not going to present it in this order, because, well, this is a case study, and case study is a study of a project that happened through time. So we're actually going to switch over to a little bit of a timeline here and walk you through the Britten project timeline.
All right. So taking us back to our main objectives, which hopefully is what you hope to get out of this course today. But this is a case study, so it happens through time. We decided it would be better to tell this story as a timeline.
So here's the Britten project timeline. And we're going to have four main sections in this presentation outlined here. The first section is about business objectives-- what business objectives can generative design solve, and how do we go about identifying those for our clients and for Britten? Second section is the generative design workshop. So we actually offer a workshop, and this is where we introduced generative design tools and processes to Britten.
The third section is application. This is where Britten took what they learned and ran with it, resulting in the fourth section-- the outcomes that we came up, which, incidentally, met the business objectives we started with at the beginning. So it's a nice timeline that completes all of the objectives. And these are our four main sections.
So we're going to start right away with the first section on business objectives. And this part of the process, these are some possible business objectives that generative design can affect, things like light weighting, minimizing mass while maintaining performance, performance impact-- this talks about optimizing durability, eliminating areas of weaknesses-- sustainability-- maybe choosing different materials or production methods that cause less waste-- and part consolidation, a really cool one where you take maybe an assembly of multiple components and you replace it with a single component that does the same job the same way or maybe even better. Isaac, would you say that our project hit any of these objectives?
ISAAC THELANDER: I think so. I think that was the goal, and it kind of was a perfect fit for what it does. I don't have too much more to add there, Ryan. It's good.
RYAN JOHNSON: It's good. Isaac's excited about this, if you can't tell. Nice. So these are general business objectives that you might be aware for your own use cases.
Now, of course, for Britten we had specific objectives that we identified. So this whole process started with what we at Team D3 call the key results program. We actually had some consultants meet with Britten and identify some key results that they were hoping their company would achieve. Now, some of the details have been redacted here, but in general you can see the kind of stuff that they're talking about.
What key result would they like to accomplish? How would they measure that result? Maybe a starting value, a baseline, and a success value. What solutions does Team D3 or Autodesk offer that could meet those needs? And path to success is really where we outline in detail how we might get to that objective.
And so, through this process, we identified several things to work with Britten on over a long period of time. But one that we could tackle right away was this first one, business growth. So the key result, growing the business. The measured value, we did identify a baseline and a success value for this being successful.
The solutions were generative design training and possibly a workshop. The path to success included things like creating new intellectual property by replacing old cast ductile iron parts with new, better, modern versions, iterating quickly through product development to get quicker results, quicker production, and, ultimately, if possible, even come up with a new design that was patentable or at least marketable. So we identified all of these things in our program.
This was actually prior to Isaac joining me, but he was pretty quick to the next step, which was our initial investigation. So we've now identified we want to do a generative design workshop with Britten to help meet this goal. The first step is investigation.
I asked Britten to bring three or four possible projects that they think might be a fit for generative design. We would walk through these projects, get an overview of them, and decide which one to focus on for our workshop. And so, we see a couple of options-- BannerSaver, the banner drop, gas shock-assisted doors, and a parametric display base. In the end, we decided really that the BannerSaver was an awesome application for generative design. And that's what we went with.
The next step for me in preparing for this was just gathering some data, getting some model files from Britten, getting notes on the performance, the conditions this thing operated under, and what it needed to perform under. And so, we gathered all that stuff up, and we were ready for our next main step, the second step in our timeline, the generative design workshop.
So in this phase, I'm going to actually walk through the steps of the workshop. We are not going to be teaching you how to do generative design, because this is just a case study. We want to focus on the story. But I will do an overview of the steps in the process for you and what those iterations looked like inside of a workshop.
Afterwards, when we get to the next step in our timeline, Isaac will walk you through his own application of that. All right. So generative design workshop. Let's talk about the steps for generative design first of all.
In general, I would break down generative design process into something like this-- the setup, which I think is a cycle. You'll see why in a minute. Setup involves editing the model, creating design conditions, constraints, loads, things like that, setting up design criteria, like maybe manufacturing methods or what problems you're trying to solve, and then generating the model. Once you've done that, we're left to evaluate the many outcomes that generative design created for us.
And from that point, oftentimes we go, based on the results, back into setup mode, make some changes, run something a little bit differently in the study, and get back to that evaluation point. So we followed the circle around, we've gotten to evaluation, and maybe now we have a model we can move forward with and create. So that's exactly what we did in this workshop.
We're going to walk through the first setup with Britten. First step was preparing the model. I decided initially to utilize the existing model exactly-- just determine areas where loading conditions were needed, where things would attach, and remove everything else. So ultimately, really, we were just running the study on this single component to see how to optimize light weight and make it more efficient.
So here our are main constraint and loading areas we isolated out of that model. And then, we actually use the model itself, as well as some additional geometry, as obstacle geometry. These are areas you don't want material to be produced. For example, in this case, the banner arm comes out of this little cylinder on the bottom, so we need an obstacle area that avoided that.
Next step-- adding design conditions. So from this model, just to give you an idea, I've outlined some of the constraints and loads that we added. For this simple first iteration, I just did a simple fixed constraint on the bearing face. In reality, the bracket kind of moves around in the wind a little bit. But the idea was just to get a simple first result.
Number 2 is the wind load, number 3 is the banner weight, and number 4 is the spring load. And then, I also added some gravity, even though minimal. It's automatically included, so I just left it there.
And then we actually run three load cases at once here. So based on the data I was given, I picked 5 miles per hour, 25, and 50 miles per hour as the wind conditions. So in generative design, we set up multiple loading conditions. And the end result will be able to handle all of those.
For design criteria, we just picked a mix of options. We did have an idea what we wanted to use and what manufacturing methods we might be looking for, but it's usually good to throw in some surprises and see what the software comes up with. Yeah. And then we hit Generate.
So what did we get? We actually got a pile of results, and we're going to dig through them and tell you what we learned about them. But first I want to talk about the Explore workspace. This is one of our main objectives.
The Explore workspace, I think, is one of the most important parts of generative design. It actually allows you to dig through the options you're given and figure out which one or which ones meet your needs. And so, in this view that we're looking at, we've got some filters showing on the left here, where we can actually isolate, for example, a specific type of manufacturing or specific type of materials. We can look at which models actually have healthy converged solutions, things like that.
And the second one I want to highlight in the Explore workspace-- this is my favorite actually-- the scatterplot. We can build our own chart to show where the different outcomes land. In this case, I'm comparing mass with safety factor. So the higher on the chart, the higher the safety factor, and the further to the left, the less the mass is.
So ideally something in that bottom corner is going to be pretty efficient. We don't really need something with a safety factor of 30. So those are probably a little bit overdone.
And then, you can move around and point to different outcomes and preview what they look like on the side with some additional details. In this case, this is a converged result made out of nylon 66 with an additive manufacturing type in the Z-plus direction. So I don't know. Isaac, you probably didn't use that Explore workspace at all.
ISAAC THELANDER: Ryan, you gotta stop lying to people. I love the Explore workspace. It's fantastic for me. I'm able to create simulations to a certain extent of initial forces. And as I'd like to show you much later in the program, I use that for some completely unrelated designs just to come up with initial ideas. You can take a look at generalized where it's showing the paths flowing around whatever you need it to move around in obstacle geometry, and you can build out your ideas from there-- use this as an inspiration to go from.
RYAN JOHNSON: Awesome. All right, I'll try to stop lying about you. We will go back to-- let's look at our first outcome. So we dug through these explore options. Out of that, what I discovered was my starting point wasn't really great.
Now, I specifically wanted to start with the existing model and not create my own new geometry, because we're trying to apply it to their design. But I discovered we had a very high percentage of failed solutions, or nonconverged solutions, and a lot that just visually they didn't look great. Those bottom corners of the piece were pretty unnecessary. Overall, the geometry just was a little too complicated.
So time to run through our loop again. And we're going to take a look at my second setup. Now, this is all prior to the workshop delivery. I'm going through this process on my own. After which, we delivered to Britten in which we taught them how to do this and walked them through the same process you're seeing right now.
So here's my redesign. Not terribly different, but definitely a more simplified shape with simplified forces and connections. And we ran through that exact same process. We'll jump straight to our second outcomes here.
And so, what I learned from that-- we definitely had improved convergence on these models. Much better resolutions. They were visually superior. And even came up with some ideas that we weren't expecting, maybe looking at a design and thinking of a way we might produce it outside of the production methods that we chose.
So it was pretty interesting to see those outcomes. In the final save for the workshop, we did discover that we could remove a lot of volume from that original design, even when using a less strong, maybe plastic material. In addition, identifying possibly new options for manufacturing, like some shapes kind of looked like stamped sheet metal pieces. And not that that's where we're going to go in the end, but really it taught us that we could think outside of the box.
And it taught Isaac that generative design is pretty awesome.
ISAAC THELANDER: Yes, it is.
RYAN JOHNSON: At least I hope. Yeah. Awesome. So that was our Generative Design Workshop. Now, we're moving on to the third main part of our timeline, which is Britten starting over from scratch. Take it away, Isaac.
ISAAC THELANDER: So Ryan showed us the Generative Design Workshop. I was very excited going into it. I had no idea what generative design was starting out. But I had all sorts of thoughts. Maybe it's AI integration into design or whatever.
And I came out of it very excited as well. Not for the same reasons, but just because I had this cool new tool, this exciting new tool. They say every tool is a hammer if you try hard enough, and I kind of hammered with generative design on everything that I could get my hands on from then on.
But to focus on what we were actually intending to use generative design for, we wanted to make a new design for BannerSaver. It's one of our most standard products here at Britten. A lot of stuff we do is very custom signage, unique experiences for clients. But BannerSaver is pretty standard. It stays the same.
RYAN JOHNSON: Isaac, can I interrupt for a second? You kind of skipped over-- maybe you could describe what BannerSaver even does.
ISAAC THELANDER: Oh, yeah. Yeah. We did skip that. BannerSaver, if you've ever walked down a street or driven down a road, and you've seen those two-foot, three-foot strip banners hanging from a flagpole, hanging from the side of a building, those are probably being held on by BannerSaver. The reason being that typically those would just be held on by poles, brackets that are just clamped into the wall. A big windstorm comes, shreds the banner, sometimes breaks the poles, rips a lot of stuff all to shreds.
The BannerSaver, the idea behind that, is that when that windstorm would come, the whole mechanism would be on of a shoulder. So it would just flow with the wind on the spring axis. And it would save the banner. It would save the damage. It would save money.
So we went to go make a new version of this. We kind of, in the workshop, looked at ways to improve the version that we already had, but we wanted to look a little bit bigger than that, make a design from scratch, maybe something that we could patent, something that we could market at least as a BannerSaver 2.0 or a BannerSaver 10,000. The trouble with that, the design is already pretty good. I'm told it's based off of a human shoulder, a little bit of ball and socket joint.
So trying to get away from that initially good design to get to something even better was challenging. You can't reinvent a mousetrap. So we had to get away from just that lever action idea and break down to our basic needs. What does this have to do? It has to hold a banner-- well, it has to display advertisement. But the easiest way is it has to hold a banner.
It's got to be easy enough to make. It's profitable it's cheap for the customer as well. Everybody wins if it's easy to make. It has to relieve under wind pressure. It's got to be safe for the banner, for the building, for the pole-- whatever it's attached to. And it has to be unique enough for us to patent or at least market as our own. This is our unique BannerSaver. I'm sure a lot of you guys have the same concerns when you're working through stuff.
So brainstorming for us probably took the largest part of the design time. Generative design, I love it. I absolutely love it. You'll see that throughout this presentation. Ryan heckles me about it a little bit, but I love generative design.
But it doesn't give you ideas. It doesn't give you creativity. That's what engineers are for, or design people who give engineers headaches designing things that can't be made. But brainstorming, we spent-- me and the other engineers-- two weeks, a month, just staring at a whiteboard, when a thought came, just scribbling it down. I have a degree in bioengineering, so I tried to think of what biomimetic things I could come up with-- what things that look like stuff from the wild.
Could I do something like a tree limb? Could I do something like an elephant proboscis-- elephant trunk? Could I do something based off of the elastic structure of a chemical structure, like an elastin or lignin or even chitin? Which didn't work out, but it's a cool idea. I'd love to explore it more.
You're just looking for some sort of really novel approach, some simple approach, and then being able to use generative design to iterate and improve on it. Came up with some wild ideas, one of which we ended up prototyping a little bit. And I'm sad that it didn't quite work out.
RYAN JOHNSON: Yeah, what do we-- I had to grab a screenshot of this meeting. What are we looking at? What in the world is this?
ISAAC THELANDER: It's my beautiful brain baby. Well, not only mine. It's modeled after an elephant proboscis or a octopus tentacle more like. And it's not seen in the wild outside of some very expensive surgical robots. And the idea is that-- I don't know if I can really-- I can't point at the screen.
But it stands up straight due to pressure on the ligature onto those discs in the center, kind of like a spinal cord. In this case, the spines are not strong enough or not big enough, so it doesn't stand up straight. That poor BannerSaver.
RYAN JOHNSON: So this is not the design you went with.
ISAAC THELANDER: It is not the design I went with.
RYAN JOHNSON: Well, maybe let's see what you did go with.
ISAAC THELANDER: In the end, after doing crazy ideas that we knew were crazy but were fun, we found a much, much simpler version. It's just a spring-- a vertical spring-- against an inclined plane. The great thing about that is we can tune that to whatever kind of pressure that we want to define, what kind of difficulty or wind pressure we need for it to swing x amount of degrees.
I'm not sure if it's totally patentable, just because it is so simple. I hope so. We're talking to somebody about it. It is marketable though. BannerSaver 2.0, BannerSaver 10,000-- it's much easier to produce. It's simple geometry. I love it.
They say a true engineer's job is done when they've removed everything in excess from a part, and I think we've done this. Just an axle, a spring, and an incline plane. And here we have the generative design that we whipped up for it or that I whipped up for it. It's just four parts.
It's a free-body diagram in space in three parts, and an obstacle geometry, and some forces. It's like the most pure form of engineering. Here is exactly what needs to be here, and the computer can do the rest.
RYAN JOHNSON: Nice. Nice. I missed a cue there. I kind of thought you were moving me to the next slide, but you were just talking about how simple and elegant this thing is.
ISAAC THELANDER: I love generative design.
RYAN JOHNSON: Do we want to see what generative design made from this?
ISAAC THELANDER: Yes, please, Ryan. Please. [LAUGHS]
RYAN JOHNSON: Let's do it.
ISAAC THELANDER: And then we had the outcomes page that we discussed earlier-- Ryan brought up. I used just about every manufacturing option, every material option I could, just because, in this very first design, I wanted to see everything that the computer could tell me-- learn as much as possible from it. And in that, a lot of it ended up looking very similar. You can kind of see a piece of the Explore page here, where it's all the same triangle structure with some sort of connection, single or double structured, between those two top pieces.
There's a lot of designs that we found. I ended up trying to test out the ones that were based on the injection-molded nylon design from generative design, because that's the direction we wanted to go-- a lot cheaper, a lot lighter, a little bit bigger than some of the other outcomes that we saw in the Explore workspace. But it made sense. I just love this.
Ryan, it says it didn't save any time at all on the screen.
RYAN JOHNSON: Yeah. I thought that's what you said.
ISAAC THELANDER: That's not what I said. I-- I love this. It's-- and again, Ryan does this because he wants to get me to gush about it. But I do! We set it up in generative design, hit play, and it made this design for me from scratch. I could import this into my workspace. I 3D-printed a model straight from this.
It saved me hours. It saved me days of work building and modeling this on my own, scraping away everything that I didn't need. Whereas I just had to put in three things-- simple things-- and I was done. It's fantastic.
RYAN JOHNSON: That's awesome. What did you do with these-- actually, we're going to look at some of your 3D. What did you do with this?
ISAAC THELANDER: This particular part of the case study is a great example of user error. You can see that on one face of this design, there's this ring structure that reinforces it against some moment forces. On one of these faces I'd put a moment force. On the other face, I did not put a moment force. The computer, seeing that I, the engineer, was true and wise and knew everything, did exactly what it was told. And what it was told was wrong.
When it came to destructive testing-- I misplaced it. It's in a photo. But it broke. It broke, and this is kind of an example of the iteration that we started to go through.
There's a few iterations that I don't show the in-betweens of, but it's making a design, 3D-printing the design or putting it into a model space, seeing what works, what doesn't work, and iterating from there, mostly 3D-printing and destructive testing. Avid prototype destroyer I am. So we made incremental improvements from BannerSaver 2.0, BannerSaver 2.1, all the way up to BannerSaver 10,000. Not 10,000 improvements, but yeah, just iterative improvements to both the pole mount and the base.
RYAN JOHNSON: Yeah. I see you're extending a little bit on our work in the workshop here. I solely focused on the bracket piece itself, but not the part that held it on. And now, you're taking that and running all of these pieces through it. So that's pretty awesome.
It's a surprise to me that we're hearing the word destruction again on one of your slides.
ISAAC THELANDER: Yeah. I have themes-- excitement and destruction. I sound like an adult toddler.
RYAN JOHNSON: I was trying to get a slow motion video of one of these breaking into pieces for the presentation, but--
ISAAC THELANDER: Right.
RYAN JOHNSON: Maybe if anybody is there live, we'll have one for--
ISAAC THELANDER: Right.
RYAN JOHNSON: --the live session.
ISAAC THELANDER: Yeah. If we can smuggle one of these through airport security, we'll break it live. You can see some of the iterations here that we went through and some of the broken versions of the iterations showing, where new factors needed to be added to the generative design study, or just some bulk added by me manually. It's not as pretty as generative design, but it's fast.
RYAN JOHNSON: Yeah. I love that. A lot of times, in my experience, people maybe don't totally trust the results from generative. I think what I heard you saying is that some of the failures you discovered were due to your own miscalculation of forces or the way that you put the forces on. So I think it's important to note that you can go through many iterations with generative design.
Yes, we advertise it as being this thing you run once, and get all these options, but you don't have to stop there. You can work off of the results you gathered and keep improving with each cycle. Pretty awesome.
Cool. Well, thank you, Isaac. It was really fun watching all of the stuff you did with this. Our last step in the process was to just zoom out and look at the outcomes. What outcomes did we reach?
Before we do that-- and look we're going to look at BannerSaver 2.1, but I want to zoom back out to the very beginning of our presentation. Just to remind everybody, this started with us identifying a key result that we wanted to accomplish and figuring out how we were going to accomplish it. This led us to the Generative Design Workshop, which led Isaac then to continue iterations on top of that.
And where did we end up? Isaac, tell us about the current version.
ISAAC THELANDER: Well, I'm very happy with the way that this has helped us. Our goal was to make something that was sleeker, was cheaper, would make our company more money in the end-- more cost effective. And I think we did. Going from this big cast piece to this small, [? K&B ?] injection-molded design 3D-printed here. Going from 14 pounds of weight down to 3 pounds of weight. The material is cheaper, the process is cheaper and faster to produce in long time or big runs. The linear force deflection curve-- one of the only things I disliked about the original design-- I was able to improve upon and make it a little bit safer for the banner even.
It's a better design. It's prettier from a marketing standpoint. It's a win all the way around.
RYAN JOHNSON: That's awesome. So too bad you're done with this now, and you probably will never use generative design for another project ever again. Oh, wait. We have another slide.
ISAAC THELANDER: Oh, no. [LAUGHS]
RYAN JOHNSON: So I have talked with you a little bit about what other projects you think you might be able to use this. Now, I know these are all just in ideation stage right now, but maybe you could tell us a little bit about some of these. Here's one.
ISAAC THELANDER: Yep. This was a custom design for a client. They wanted a big, hanging, rotating atrium sign. So this was one of my first projects that I tried generative design out on. That's why it's a little more rough. But I was able to go from generative design, from the three hanging points and the external ring that the client wanted, and build a very simple, easily-weldable tube frame roughly based on that design. It was exciting.
RYAN JOHNSON: Yeah. I remember you talking about how you were maybe surprised sometimes that the choices IT made or the directions and [INAUDIBLE] that it added, which is pretty awesome. Even if you're not directly using this model, that you can use it to inform other types of production.
ISAAC THELANDER: Exactly. Exactly.
RYAN JOHNSON: Yeah. That's cool. Here's another one. What in the world is this?
ISAAC THELANDER: This was a custom marketing display for things like a BannerSaver or something that hangs off the edge there. And for this, I think it's an example of how you can save a lot of time with generative design. I made the base. I made much more base than was necessary, just so that it could narrow itself down to what it specifically required.
I made some obstacle geometry and some [? keep ?] geometry. And I got a bunch of results that informed me how I could build this frame. And it's nothing near what I would have built otherwise. It's much more organic, much more creative, and great for marketing.
RYAN JOHNSON: There you go. Marketing. All right. Now, this next one, this doesn't really look like what I would think generative design could be used for at all. What in the world is this?
ISAAC THELANDER: Right. When you have a hammer, every problem looks like a nail. And when you have generative design, every problem looks like something that can be solved with generative design. In this case, this was a box pop structure from our shipping container display and experience division.
This is a bar for a client. It's two 40-foot containers laid side by side, connected together, and with a deck for people to enjoy themselves up top with. I wanted to make the simplest, cheapest way to support this and still be pretty safe. So I threw the whole thing into generative design, upped the safety factor a million degrees, and hit go and saw where it would inform me that it needed support.
And the results were always very consistent. And went from this straight to a design. In this case, again, a very simple tube frame but based on the organic help from generative design.
RYAN JOHNSON: Awesome, man. It is pretty cool seeing some of the stuff you're doing with this. Thanks for sharing all that.
ISAAC THELANDER: Thanks, Ryan. I appreciate it. I appreciate you.
RYAN JOHNSON: Awesome. So we are now at the end. We just hit the project outcomes. So just to recap, we started with business objectives. We went through a workshop phase, where we taught Isaac and Britten about generative design. And then, I think, crucially, what they've done that a lot of clients haven't done is took off and run with it themselves. And Isaac, you're doing some awesome work, resulting in the outcomes that we saw, some of which, again, just to close the loop, tie directly back to the business objectives that we identified at the outset.
So it was a lot of fun. Thanks for working with me through it, Isaac. Hopefully everybody got these four main points out of our little presentation, our case study here, looking at business objectives you can solve, watching iterations of a generative design study, using the Explore tool to evaluate all of those outcomes, and looking at some real-world outcomes. So hope you all enjoyed that. Thanks for joining us, and enjoy the rest of Autodesk University 2023.