CAD storyboarding

This is something I’ve done in the past and wondered if anyone else works this way.
Context scenario: sketches are bought-off, and its time to start doing some 3D modeling in SolidWorks.
While doing sketches the designer often thinks about ‘how’ this will be built, but now its time for the mouse rubber to hit the road.
First task is to create a ‘storyboard’, similar to those used in film, to sketch the ‘build’ process…e.g. first, make a plane in the right angle/direction. Then make this perimeter curve. Then a surface, and offset that. Each feature, or group of features, is manifest in a storyboard panel.

I find it helps to ‘think through’ a build beforehand. I’ve heard of Alias techs sketching out their meshes and patterns, how to build things in an efficient, elegant, and updateable way.

I will try to find some past examples to post up here.

Yikes, seems like twice the work.

It sort of is, but only in the way that planning a project of any kind - versus jumping right in - is twice the work. A man with some wood and a hammer and nails could probably just go and build a house, but for various reasons might want some drawings or blueprints first, and explicit work orders if he’s inexperienced.

This is sort of what I’m talking about, but is meant for Alias patch layouts.

[ Deleted ]

We are very close to this now. I worked at a sheet metal factory in the past. While there is currently still a human involved at each stage, the system ultimately takes each 3D part, that has been continuously simulated throughout modelling to check that it is possible, and outputs a full bending machine program, punching machine program, accurate flat patterns and the scannable barcodes for the operator. It will be very impressive when this can be done for more complex tasks like IM.

On topic, I like the idea. I have written down how I might model complicated parts before, although I found that ultimately, to storyboard a cad plan in advance you need the experience to know what steps to take. By already knowing what steps to take you know, roughly, how to model the part anyway, and therefore don’t really need the plan. I think it would work if you were handing the cad off to someone less experienced so that they can follow guidelines and model without wasting time.

Wouldn’t necessarily go step by step but in Alias I’d do a decent amount of preplanning for late stage surfaces where I wanted to maintain construction history (the ability for surfaces to update is very finnicky and if you don’t modify the correct input it explodes anyways).

By thinking through that order of operations you could make sure that you had enough room to tweak and experiment without having to redo dozens of trim and untrim operations.

I do this to a certain extent. If I’m dealing with a complicated CAD problem I won’t try to immediately jump in and hammer away at it. I’ll work through different approaches on pen and paper. In my experience if you don’t know where you’re going you’ll end up rebuilding the CAD multiple times and it’s a very frustrating and slow process. I don’t go down to a literal step by step though, I keep it fairly high level.

Ultimately, I think this kind of figuring out is pretty common across problem solving in general. It reminds me of pseudocode in computer science. When facing a complicated problem, programmers will often write code in a human readable form and mentally work through what the computer is doing at every step. I remember as a kid having my engineer dad walk me through how we were going to build the roof for our workshop by chicken scratching a little plan. I’ve used the same technique with skilled craftspeople to work though how we were going to build complex steel structures. Seems like the same idea just in digital form.

That is true Lois, it is “the process” for creative problem solving.

  • Low fidelity concepting (call it a sketch, a mockup, fail fast, whatever)
    Analyze (look at it objectively, maybe get feedback, think about it)
    Iterate (make adjustments, refinements, more concepts)
    Prototype (make some stuff whether they be refined sketches, CAD models, bread boards, whatever)
    Analyze again (look at it objectively, maybe get feedback, think about it)
    Repeat as needed

I’m definitely not the best in CAD, though I have a few go to contractors I use who are amazing. Learning it has helped me to at least empathize with them, understand what they need to get to the desired output, and help them to strategize a build like this.

Sean, I think it is good to post this so that other designers know it doesn’t just come out of thin air. Like any skill, CAD is something that needs to be thought about and worked at. I’ve got a friend who was an early Alias user. He used to drive around with an SGI machine in his back seat going from client to client. Now he is a 3D designer at Hyundai/Genesis. They don’t call them modelers, or CAD jockeys because often times he is getting an approved sketch that is super conceptual and he actually works through how that will become a design. It reminds me just how big design is. There is room for all different kinds of designers, so many flavors of specialists, generalists, and hybrids… sorry off topic. A little pontificating as procrastination.

That is a fascinating idea. You have to build ‘intelligence’ into a SolidWorks parametric model, but what if the CAD knew something about what it was you were making, or could have ‘learned’ from prior models? Then the designer could ‘design’ by storyboarding the build, and course-correct along the way, and hopefully just nudge some points in the end to make a viable concept. Seems more on the designer’s side than the crazy generative Voronoi goop that gets passed off as AI-CAD.

Right, step-by-step would not be necessary, and the # of steps might decrease with a user’s proficiency. Sounds like you are/were pretty sharp in Alias. “Thinking through the order of operations” is the key sentence and idea. Thanks.

Yes, and the converse could be that a less experienced CAD designers can take their storyboard and review it with a director who could advise on how to save time or get a better result. Its true that you need to know the steps to even start knowing how to plan, but younger designers are going to take one tool and use the hell out of it (“to a man with a hammer everything looks like a nail”) where there could be more appropriate tools.

And I bet those “go-to” contractors would be able to whiteboard sketch their process ahead of time, building the product before your eyes, having a highly detailed mental model of your product in their heads and then being able to communicate it to you. Then you can make adjustments as needed to the build process for where you might want to enact later changes… so being able to think through the order of Ops is both a good mental discipline, and a method to communicate to other interested parties.

Your Hyundai friend might be able to confirm/deny this, but I’ve heard auto studios use different staff for even conceptual sketching vs modeling. The designer will sit with the modeler who will attempt to enact what the designer sketched, but tempered with better knowledge of good surfaces, materials, reflections, etc. … and if the designer is bringing up some really trick new styling element the CAD designer might need to storyboard the steps to achieve that result.

I’m fairly proficient in SolidWorks but still think it mostly sucks as a design tool. Ergo needing to make storyboards, and urging my team mates to do the same, lest they go wandering in the feature tree desert.

What Louis said.

Step-by-step planning sounds like post-its and discussions to me. You can skip it and do a lot by just brain power.
Don’t create discussions on your WIP, but discuss as soon as you have a working version.

I do lay-out certain plans (in case it is still hard to do by using only the brain) but mostly by sketching a few views as thinking aids. It is very important in case you want to branch off different versions with different features, or if you have complex surfaces and need to figure out how certain blends and fillets will come out. The devil is in the details - you can’t always see him beforehand and he might just destroy that nice idea, so you’d have to start over. And unless you’re at the beginning of learning CAD I don’t recommend starting over too often.

The AI options are interesting, definitely. As a next step I want to explore 3D sketching in an AR space. From there the AI could suggest modeling options.