In this second article of a four part series, Thomas Teger, CPO of swatchbook, takes a look at how materials are being used and represented in the digital design process and how, once again, design and engineering are seriously disconnected.
Materials are at the core of everything
The materials on any product define how a product looks, how it is made, how it feels. They determine whether a product is “expensive” or “cheap”. Colors, patterns, textures, finishes of materials let us make an emotional connection with products within nanoseconds. Materials combined with craftsmanship will determine how long a product will last. Materials combined with colors and graphics help companies sell the product inside.
Materials have evolved
Materials of any kind – metals, plastics, paints, fabrics ….you name it – in every industry have evolved significantly over time, including manufacturing, processing, and recycling. Harder, stronger, lighter, with minimum amount of environmental impact, scientists around the globe are advancing the science of materials on a daily basis.
And while it is easy to understand materials – their behavior, structure, texture, weight – by just looking at and touching them, the fundamental question remains how to best represent them in digital form.
Materials are incredibly complex, or at least they can be. From its chemical compositions and properties, supply, maintenance and repair, all the way to equivalents and alternatives, the information goes incredibly deep. And there can be thousands of them. And then there are composite materials that make it even more complex.
It’s complicated, for sure. So the question is how are, can and should materials be represented in digital form?
Digital representation of materials in design and engineering; disconnected, again
When we take a look at the digital product development process, we find that the use of materials is split between designers on one hand, and engineering and manufacturing on the other. The different parties in the process have very different needs. While there is certainly overlap, designers care mostly about the aesthetics of the material which includes anything from material type to color to finish, while engineering cares mostly about the engineering, sourcing and manufacturability of the material, and how selected materials will perform on the product.
Even though the needs are different, the information needed by each involved party is requested from the same material. So, in a perfect world, there would be a single representation of a material that contains all of that information, managed in a single system that anyone can access to obtain the information they need at any given time.
This brings us back to last month’s article where we discussed the disconnect between designers and engineers when it comes to modeling and PLM solutions. The problem with materials is even bigger. Designers and engineers have their own representation of materials. And the only connection between the two representations is the name, and color if you are lucky.
This is a fundamental problem with the systems that are being used. While you can always get your geometry data from a modeling application the designer is using into a CAD system using geometry standards, or mesh representations in a worst case scenario, or going from a CAD system to a visualization application, there is really no way of exchanging material information between systems. The best you may get is color information, and, depending on the file format you are using for exchange, you may even get a material name. But that’s it.
This also goes the other way when transferring data from a CAD or modeling system into a visualization application.
CAD applications that offer dedicated tools for designers for modeling and rendering in the same environment – not that they are widely adopted due to the non-pleasant user experience as we’ve discussed – provide material libraries. Yes, libraries! Multiple! One for designers, and one for engineers. And they have NOTHING to do with one another. The one for designers is purely for visual purposes, while the one for engineers contains actual material standards, but no visual representation.
From a management standpoint engineering and manufacturing are using PLM systems, which are powerful as we learned last month, yet often horribly non-user friendly applications conceived by engineers for engineers. So, not only can these systems be atrocious to use for a designer, they also don’t carry the information that is needed for their creative tasks on hand.
How do designers manage their materials? They don’t. Well, they do, but more in a non-sophisticated way by using shared folders on the company’s network, rather than going through a dedicated system.
Materials in Computer Graphics (CG)
Let’s take a look at materials in CG. Ok, you may think, but as long as I’m in the design space and I don’t care about engineering standards I should be good, right? Not so fast.
Unfortunately there is no standard that allows for an easy transfer of digital materials from one application to another. Whether it is KeyShot, VRED, Modo, Cinema4D, Solidworks Visualize, VRAY, Maxwell, or any other rendering application, they all have their proprietary material definition. With the use of neutral file formats such as FBX or Alembic it is possible to exchange color and texture information, but the actual parameters that make the material look like metal, plastic, glass, paint, cloth, wood etc. are not being transferred.
The entertainment industry is trying to do the right thing. MaterialX “is an open standard for transfer of rich material and look-development content between applications and renderers.” FX (sound and visual effects) companies are using various visualization and rendering systems in the production of a movie, including custom developed solutions. Having to recreate materials and scenes to get a 100% match between systems is an absolute must in the production of a feature film in order to keep production times in check, as more and more effects are being added to new films.
This is certainly a much needed undertaking, and something that the DMP industry (that is the industry that designs and manufactures Discretely Manufactured Products such as planes, trains, automobiles, cell phones, furniture, toys etc.) needs to adopt as well. It certainly doesn’t have to be to the extent of what the entertainment industry is trying to do, but at least having a way of exchanging materials between the various rendering, modeling and CAD applications would be a great step in the right direction.
Good lookin’ … but no meta data
Aside from lacking a standard, what all these digital materials are lacking are meta-data. In CG, all materials are representing are the visual aspects of a material: color, finish, texture, pattern. But you don’t know anything about the actual material, its composition, color standard, kind of finish, weight, thickness etc.
For CG this is certainly more than enough. All that CG artists and visualization specialists care about is, “what does this material look like on a 3D object?” Especially when this is all you want to communicate, may it be during a design review, a design approval, documentation, or marketing imagery or behavior under stress.
But what if you want to communicate the use of the materials back to engineering and manufacturing? How do you communicate what material you used? What color has been applied using what color standard?
None of the design or visualization applications give you an easy way to capture the meta data of a material. And even if there is, then the challenge becomes how to communicate this information? Because, as we know, designers don’t use PLM.
Material libraries; nothing but meta data
There are plenty of material libraries that have been put together for engineers and manufacturers, whether as part of a CAD system, or part of a dedicated external system. While these material libraries provide deep technical information using standard names and descriptions, there is no visual information associated to the materials.
This is also what you find in PLM systems where the materials are being managed. If you are lucky, you get a thumbnail with the material. Otherwise, it is basically a glorified spreadsheet that describes the material.
How is that usable for a designer? Well, it is not. Hence the disconnect.
In fashion, designers are really struggling
As discussed, most companies that produce discretely manufactured products are using material libraries in design for visualization, for the creation of marketing materials, for engineering and manufacturing.
When we take a look at the fashion industry, the idea of digital material libraries is just now something that is being thought about. There are certainly larger companies that have figured it out, and have created material libraries for designers, yet the majority of companies are still at the beginning of trying to figure it out.
While most fashion companies are about 20 years behind in particular when it comes to 3D design, they all miraculously adopted a PLM system – something that is, unfortunately, often the source of complaint from the IT people managing the system. Designers in particular refuse to use PLM for reasons discussed in last month’s article. While these systems are incredibly useful for manufacturing, sourcing, managing inventory etcetera, the information contained for a particular material is of minimal use for a designer. At best you get a thumbnail view of the material, which may give you a visual reference but nothing really useful for the design process.
Another challenge in fashion is that you are dealing with potentially hundreds of new materials (fabrics) from various suppliers, each season! And depending of the size of the brand, this can potentially include hundreds of suppliers. Whether these are new fabrics created by suppliers, or requested by the fashion brand, the way these materials are being shared and communicated are either by spreadsheet, or by sending physical samples back and forth between the fashion brand and suppliers.
There are certainly standard fabrics that are being reused over and over again, but even those may have different colors or patterns based on the season.
Other industries do this as well, but on a much smaller scale. Take automotive paints for example. Coating companies such as Axalta come up with new paints and new formulas every year. However, all these paints have a much longer “shelf live” than fabrics – which only lasts one season. And an automotive company is not sourcing paint from potentially hundreds of suppliers but only one, or possibly a few coating suppliers depending on the particular vehicle line.
When sending all of these fabric samples back and forth, and creating physical samples from these to visualize the fabric in context, just imagine how much gets thrown away. For example, a sportswear manufacturer creates samples for 100,000 SKUs that amount to filling up containers when stacked will equal the size of 4(!) One World towers. That’s a lot of waste!
And then there is the communication challenge. Whether it is internally trying to make design decisions while putting together a digital 2D tech-pack and a physical fabric sample in hand in your head, which results in meeting after meeting after meeting, or, once the decision has been made, with manufacturers who are tasked to produce the new design with a certain fabric. Now you need to send the 2D digital design via e-mail, and the 3D design via FedEx, and make sure you don’t forget to send one but not the other.
In fashion, the behavior of a fabric is an integral part in making decisions when designing a new garment. While there are systems out there that capture that behavior, fabric simulation doesn’t adhere to any standards, and require again specialized software to capture and use this data. And PLM systems don’t store this information either.
Moving to digital material library, or bringing these various libraries together, will do so many things, except for one: it will not replace touch and feel. But what it will do is reduce the dependency on physical material samples and prototypes, and allow you to make better decisions faster, which ultimately will result in better designed and manufactured products.
In the future there will need to be a standard file format for materials that contains all layers of information – visual, meta data, simulation, custom data – where the individual systems just access the information that is needed and utilize it accordingly. There are some good attempts being made, for example by Nvidia with MDL, but it doesn’t go far enough, and it is not open-source which would allow it to be driven by the community rather than Nvidia’s agenda.
There are a large number of participants in the digital development process that need access to materials, but as we discovered the information that is available is either not what they need, or it doesn’t tell them anything, or it is tied to specialized system.
Being able to visualize materials at any point in the process with the right information at hand, and then being able to visualize them easily in 3D without having request all physical samples, will allow designers to make better decisions quicker, and spend more time focusing on their actual design task than on mastering complex systems.
This is not a threat against PLM. It should be seen as a complementary solution that caters to the specific needs of designers.