CMF decisions tend to multiply quickly. A new seat textile changes how the surrounding plastics feel, which changes the trim, which might change the headliner. Even when the vehicle itself is fairly resolved, there can still be dozens of material directions worth exploring.
The problem is that creating every one of those directions in 3D takes time. If we want to test a slightly different textile, plastic grain, or surface finish, we either need to find something close enough in an existing material library or stop designing and build the material ourselves. That's the part of digital CMF we've been trying to make much faster with twirl.
Instead of starting with an existing material and modifying it into what we want, we can describe the material and generate a complete seamless PBR material from it. That makes it much easier to explore several CMF directions on the actual vehicle before spending time developing physical samples or production-ready digital materials.
Digital CMF gives us more room to explore
CMF stands for color, material, and finish. In transportation design, it covers everything from textiles and polymers to leather, metals, coatings, paint, and decorative finishes. Digital CMF brings those decisions onto the 3D vehicle so we can apply materials directly to the model, change them quickly, and evaluate several directions under the same lighting and geometry.
Porsche has presented a digital color and material process that includes more than 900 digital materials alongside calibrated scenes, displays, and virtual and physical light booths. Autodesk's overview of Porsche's workflow is a useful example of how far this can go when the goal is reliable digital evaluation rather than a quick concept render.
That level of calibration matters when a digital material is being used to represent a real physical material accurately. Earlier in the process, we're often trying to answer a simpler question: what material do we actually want?
Start with the material you want, not the material you can find
Most material libraries work in the same direction. We search through what already exists until we find something close enough to the surface we have in mind. That works well when the material already exists. It becomes limiting when we're exploring something specific.
Maybe we want a warm gray polymer with an unusually fine grain. Maybe we're looking for a recycled textile with larger natural fibers. Maybe we want an anodized metal somewhere between champagne and silver with a fine bead-blasted finish. With twirl, those descriptions can be the starting point rather than the end of a long material search.
twirl generates the description as a seamless PBR texture set, so we can put it directly onto the vehicle rather than trying to recreate the idea from a generic plastic texture. If it isn't right, we can change the description and generate another direction. For CMF exploration, that ability to iterate is where the workflow becomes useful.
PBR materials let us evaluate more than color
Two surfaces can have nearly identical colors and feel completely different once they're under automotive lighting. Roughness changes the reflection, normal information changes how the surface breaks that reflection apart, and the scale of those details changes how we perceive the material.
twirl generates these properties as a complete PBR material rather than giving us a single reference image. We can put a generated textile across a seat, use a generated polymer on the IP, and compare both under the same environment. If you're new to the format, our guide to PBR maps explains what each map is doing. You can also browse the twirl PBR material library when an existing material already fits the direction.
Generate several directions before committing to one
Instead of spending an hour developing one plastic material, we can generate several plausible directions and compare them on the same model. We might start with a fine-grain charcoal polymer, generate a warmer version, try a larger grain, and then test a softer matte finish. For a seat insert, we could compare a technical knit, recycled woven textile, and synthetic suede without manually authoring each material first.
Adobe, ArtCenter, and Rivian explored a similar digital-first approach during a transportation design workshop where students developed multiple material directions using Substance 3D, including scanned physical materials and variations in gloss, translucency, pattern, and layering. Adobe's workshop recap shows how naturally digital material iteration fits into transportation design education.
twirl isn't trying to replace a tool like Substance Designer when we need a deeply controlled procedural material. It solves an earlier problem: getting enough useful material options onto the vehicle that we can decide what deserves more development.
Physical materials still matter
A generated material can't tell us whether a textile feels right, whether a polymer can be manufactured with a particular grain, or whether a supplier can meet an abrasion requirement. When we're accurately reproducing an approved material, scanning and measured material formats can make much more sense.
VRED supports this more advanced end of the workflow with Substance materials alongside formats including AxF, BTF, MaterialX, and MDL. Autodesk's VRED material documentation covers those workflows in more detail.
We don't need that investment for every idea we consider. That's the gap twirl is built around. We can generate the surface we have in mind, put it onto the actual design, and iterate until we find a CMF direction worth taking further.