In his fourth and final article for WhichPLM, Thomas Teger, CPO of swatchbook, discusses the importance of standards, how this has been addressed for 3D geometry, and why standards for materials – in particular for the apparel and footwear industry – are desperately needed.
Why do we need standards, you may ask? If everyone were to use the system from the same supplier then we would not need any standards. Everybody would be using the same native file format. This is certainly how big software companies like you to think. And companies who deploy their solutions are usually large enough so they can tell their suppliers what software they want them to use.
But the reality is that the digital process nowadays goes far beyond product development.
What about visualization for sales and marketing, for example? AR/VR/XR? Sketching? 3D web publishing? Retail? The list goes on. It is impossible for one company to do it all, or at least to do it all well. Therefore it has become, over the years, increasingly important to be able to take data from one system to another with as little loss as possible. Standards can help tremendously in this case.
A thought about cost
Transferring data from one system to another can be immensely costly. Unless the 2 systems are built to be able to read each others’ data – there are plenty of companies out there that specialize in exact data translation, and provide licenses that can be licensed – the only way to go from one system to another may be to rebuild. Or you will need to find a different solution that does support the file format that you are trying to import. Or you try to do it in the original system. No option is a viable solution in the ever increasingly demanding product release cycles and ever shrinking time-to-market deadlines. That’s why standards have been developed and are immensely important in the digital product development process.
The world is full of standards for 2D and 3D, but not for materials
When looking at 3D modeling and visualization applications we can see that there are quite a lot of standards in place already:
- Images – This is probably the most commonly used standard you may not even think about. File types like jpeg, tiff, png, and gif are “standards” for visual communication.
- Geometry standards for CAD – There are a number of geometry standards that have been defined by independent organizations that were trying to help solve an interoperability problem between software applications, rather than by companies trying to sell more software. The most popular standard geometry formats are IGES, STEP, and DXF
- Geometry standards for other applications – Over the years, additional geometry formats have been developed in particular for the use in visualization and entertainment which carry more than just geometry information, but also information for materials and textures, as well as animation. The most popular formats are OBJ, FBX, and ALEMBIC.
- Hybrid geometry and visualization standards – JT, developed by Siemens, and 3DXML, developed by Dassault Systèmes, are standards that carry a lightweight representation of the 3D model for visualization and data management.
But when it comes to materials, standards are pretty sparse. One can argue that MDL is a standard for materials, however, at its current form it doesn’t go far enough.
Materials need help
We took a look at materials and how they are being defined in the digital world in the piece we wrote back in May. We found how highly disconnected everything is. Between design and engineering, and between individual application for visualization, materials, as important as they are, adhere to no standards at all.
When looking at the fashion and footwear industry in particular, where pretty much everything is about materials, and a slow but steady movement towards 3D is underway, it becomes clear how desperately a standard for materials is needed.
It goes further than just the visual aspects
But as we have discovered, for materials, it goes – or better it needs to go – much beyond the visual aspects:
- There is also metadata including physical properties and more
- Simulation of the material, analyzing the material’s behavior under stress
- User specific data to document any additional information that is part of the material
As we have discovered in our previous articles on materials, there is currently no material format available that supports all this information. Even the available systems for modeling or rendering aren’t architected to support all of this information.
The problem with simulation
We talked about how materials have different formats depending on the rendering application that is being used. A transfer of materials from one system to another is hard to come by.
The problem is even bigger when you look at the simulation of materials. In particular when we talk about analyzing the behavior of fabrics. Each of the leading clothing design companies – Browzwear, CLO, Optitex – are providing their own proprietary fabric simulation kit. While these work great with their software, the measurements don’t translate between the applications.
This is a particular problem for suppliers. While brands have the resources both monetarily and staffing wise to acquire potentially all three fabric analyzers, where will this leave suppliers? Brands want their supplier to provide data in digital form. Having suppliers to commit to doing fabric simulation at all will be hard enough, let alone using all three of them.
This, again, is the perfect case for a standard for materials.
What about the scanning?
When talking about material standards, we certainly need to look at companies that provide hardware and software technology for capturing the visual aspects of a material. Both Vizoo and X-Rite are on the forefront of material capturing. While similar, their approach vastly differs, in particular since X-Rite is going for the “total appearance capture” (TAC) which captures all visual properties of a material instead of just the texture, transparency, and specularity of the material.
The question that comes up is what can you do with the captured data? Both file formats are different, and in particular X-Rite’s proprietary AxF file format is currently only supported be few visualization applications.
MDL to the rescue?
Nivida announced a few months ago at Siggraph 2018 that their own MDL (Material Definition Language) format is now opensource. This is a great move in the right direction. It will at least take care of the visualization aspects of the materials. But it still leaves the questions about metadata and simulation data unanswered. Yet Nvidia also has put together an open committee of brands and vendors that is in constant communication about the future of MDL. Could MDL be the standard format for materials, that even works for both Vizoo and X-Rite?
Just announced: U3M
As I’m wrapping up this article, Browzwear and VIZOO have announced their collaboration on a new open material format called U3M. According to their release this new file format combines visual and physical data in one single file format. It will be interesting to see how this plays out. A big part of this being a success is the introduction of a standard format for capturing the physical data of a material as mentioned above.
3D.RC – The 3D Retail Coalition
There are several initiatives underway to take on this task. The 3DRC – in particular the subcommittee on innovation – has taken on exactly the initiative, envisioning “highly transferable material libraries, a standardized open materials data platform and efficient, high quality, cost effective testing methods.” Just recently, (why did it take so long?), they also opened up their consortium to vendors so they can actively participate. The 3DRC has also opened up enrollment.
There is certainly a lot of movement in the industry to move towards a standard material definition, and to expand the material definition beyond the visual properties. One important aspect of defining a material standard for the industry is that software suppliers will need to support this standard. In order to do so, it will be necessary for these companies to be actively involved so they can make their voices heard. I know there is reluctance, since most developers of proprietary rendering engines take a lot of pride in the definition of their proprietary material definition. Yet the definition of a material should not be what makes a piece of software great. It should be about how this information is used by the render engine.